JP2001050848A - Design of pneumatic tire, optimization analyzer and storage medium storing optimization analyzing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance design/development efficiency of tire while predicting the tire performance being used actually with the effect of slip angle and camber angle being taken into account. SOLUTION: A tire model is generated from a tire design plan, pavement conditions are inputted by selecting the coefficient of friction μupon generation of a pavement model, initial conditions for imparting an angular speed to the tire are set and then design variables and restrictive conditions are determined (100-104). Subsequently, a model having design variables varied by a unit amount Δri is determined (106-110) and tire performance is predicted by imparting conditions for calculating deformation of the tire model and imparting a slip angle or a camber angle (112). Sensitivity is operated for each target function, restrictive conditions and design variables, a predictive value of variation in the design variables for maximizing the target function is determined while predicting the tire performance and taking into account of the restrictive conditions, and the operation is repeated until the value of the target function is converged (114-124). Shape of the tire is determined based on the design variables when the value of the target function is converged to a predicted value.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for designing a pneumatic tire, an optimization analysis device, and a storage medium storing a tire optimization analysis program. A pneumatic tire design method that enables efficient and easy design and development of the structure and shape of the tire that achieves the target, and that enables the design of the best structure and shape of the tire and the design of a tire with high cost performance. , An optimization analysis device, and a storage medium storing a tire optimization analysis program.

[0002]

2. Description of the Related Art Tire design has been based on empirical rules obtained by repeating experiments and numerical experiments using a computer. For this reason, the number of prototypes and tests required for development became enormous, the development cost increased, and the development period could not be shortened easily.

[0003] The tire design includes tire structure, shape, pattern design, and the like. The structure, shape, and pattern design of the tire include the structure, shape, pattern, manufacturing conditions, and the like for obtaining the desired tire performance. Is to seek.
The performance of the tire is a physical quantity or an actual vehicle feeling evaluation result obtained by calculation or experiment. Conventional tire design methods, such as tire structure, shape, and pattern design, have been based on trial and error empirical rules by repeating experiments and numerical experiments using a computer. For this reason,
The number of prototypes and tests required for development became enormous, the development costs increased, and the development period could not be shortened easily.

[0004] To solve this problem, mathematical programming,
Techniques for finding an optimal solution, such as an optimization method using a genetic algorithm, have been proposed. The present applicant has proposed a design method described in International Publication No. WO94 / 16877, which has already been filed, in connection with this mathematical programming.

[0005] Finding the optimal solution is likened to climbing a mountain. At this time, since the altitude of the mountain is related to the performance, the optimal solution corresponds to the top of the mountain. When the objective function is simple, the design space (mountain shape) has a peak shape with one peak, so that an optimal solution can be obtained by an optimization method based on mathematical programming.

In the development of pneumatic tires,
Tire performance to consider is actually designed and manufactured tires,
It is obtained by performing a performance test by attaching it to an automobile, and if the results of the performance test are not satisfactory, the procedure of designing and manufacturing has to be repeated. recently,
Due to the development of numerical analysis methods such as the finite element method and the computer environment, it is now possible for the computer to predict the filling state of tire internal pressure and the load state when the tire is not rolling, and to perform some performance predictions from this prediction. It has become

[0007]

However, the performance of a tire is not evaluated only when the tire is filled with internal pressure or when the tire is not rolling. For example, a tire may have a so-called slip angle or camber angle. That is, it is set from the time of mounting the tire on the vehicle,
It is generated when the vehicle is tilted or rotated when trying to turn on a curved road. Such a slip angle and a camber angle appear particularly remarkably in a motorcycle.

[0008] Tire performance is expected to be affected when the tire is tilted or rotated as described above, but these considerations are not taken into consideration, and the actual environment is assumed for the actual tire. No consideration has been given to the analysis. For this reason, at present, tire performance cannot be predicted in accordance with the actual environment, and tire development cannot be performed efficiently.

Further, in the design and development of a tire, a target value is set for a certain performance, and if this target value is cleared, the operation is tentatively terminated, and the best performance is not obtained with a given resource. Further, the design of the tire and the performance test of the tire are performed independently, and the development of the tire is performed through trial and error of trial manufacture and testing, which is very inefficient.

The present invention, in consideration of the above facts, provides a pneumatic tire design method, an optimization analysis device, and a tire capable of designing and developing a highly efficient tire while predicting the performance of a tire to be actually used. It is an object of the present invention to obtain a storage medium storing an optimization analysis program of the above.

[0011]

In order to achieve the above object, the present inventors have made various studies, and as a result, have applied an "optimization design technique" used in a different field to a special field called a tire. Focusing on doing that, we tried all kinds of studies, predicted the performance according to the actual tire behavior in the process of optimization, and made it possible to perform transient analysis, especially at the time of tire contact and rotation,
In order to streamline tire development and facilitate the provision of tires with good performance, it has been specifically established as a tire design method.

More specifically, the method of designing a pneumatic tire according to the first aspect of the present invention is a method of designing a pneumatic tire, which comprises: (a) at least including a cross-sectional shape of a tire including an internal structure and deformed by grounding; A tire model capable of giving a, a road surface model in contact with the tire model,
Initial conditions related to rolling of at least one of the tire model and the road surface model, an objective function representing a physical quantity for evaluating tire performance, design variables for determining a tire cross-sectional shape or a tire structure or a pattern shape, and a tire. A step of defining at least one of a cross-sectional shape, a tire structure, a pattern shape, a physical quantity for performance evaluation, and a tire dimension; (b) tire performance is determined by a physical quantity generated in the tire model due to deformation of the tire model. Predicting, (c) obtaining a value of a design variable that gives an optimal value of the objective function in consideration of the predicted tire performance and constraints, and (d) tying the tire based on the design variable that gives the optimal value of the objective function. Steps to design,
Each step is included.

According to a second aspect of the present invention, in the method of designing a pneumatic tire according to the first aspect, in the step (a), the angular velocity for rolling the tire model and the road surface model move. At least one of the moving speeds is determined as an initial condition.

According to a third aspect of the present invention, there is provided the pneumatic tire designing method according to the first or second aspect, wherein in the step (a), the calculation at the time of filling the internal pressure and the calculation of the load are performed, and the rotational displacement is calculated. Alternatively, a tire model to which a speed, a linear displacement, or a speed is given is determined.

According to a fourth aspect of the present invention, there is provided the first to third aspects.
The pneumatic tire designing method according to any one of the above, wherein the step (b) is: (e) a step of performing a deformation calculation of the tire model; and (f) giving a condition to the tire model. And (g) a step of giving the condition of the step (f) to the tire model after the deformation calculation in the step (e) and calculating the tire model until a predetermined time is reached (the tire model is in a transient state). ,
(H) obtaining a physical quantity generated in the tire model; and (i) predicting tire performance based on the physical quantity. The above calculation can be repeatedly performed for a predetermined time.

The invention according to claim 5 is the invention according to claims 1 to 4.
The pneumatic tire designing method according to any one of the above, wherein the tire model partially has a pattern.

The invention of claim 6 is the first to fifth aspects of the present invention.
The pneumatic tire designing method according to any one of the above, wherein the road surface model is DRY, WET, on ice, on snow,
And determining a road surface condition by selecting a friction coefficient μ representing at least one road surface condition of unpaved.

The invention of claim 7 is the first to sixth aspects of the present invention.
The pneumatic tire designing method according to any one of the above, wherein at least one of a contact area and a contact pressure of a tire model is used as the physical quantity, and a tire WET performance is predicted as the tire performance. I do.

The invention of claim 8 is the first to seventh aspects of the present invention.
The pneumatic tire designing method according to any one of the above, wherein the physical quantity is at least one of a contact area, a contact pressure, and a shearing force on at least one of an ice road surface and a snow road surface of the tire model. The method is characterized in that a tire performance on ice and snow is predicted as the tire performance.

According to a ninth aspect of the present invention, there is provided the pneumatic tire designing method according to the first aspect, wherein in the step (c), a ratio of a change amount of the objective function to a unit change amount of the design variable is set. While predicting the change amount of the design variable that gives the optimal value of the objective function while considering the constraint condition based on the sensitivity of the constraint condition, which is a ratio of the change amount of the constraint condition to the unit change amount of the function and the design variable, The value of the objective function when the design variable is changed by the amount corresponding to the predicted amount and the value of the constraint condition when the design variable is changed by the amount corresponding to the predicted amount are calculated based on the predicted value and the calculated value. And determining the value of a design variable that gives the optimum value of the objective function while considering the constraint conditions.

According to a tenth aspect of the present invention, there is provided the pneumatic tire designing method according to the first aspect, wherein the design variables are a carcass line, a folded ply line, a line representing a tire outer shape, and a line representing a tire crown shape. And a function representing the shape of at least one of the reinforcing material lines, a bead filler gauge distribution, a rubber chafer gauge distribution, a side rubber gauge distribution, a tread rubber gauge distribution, a tread base rubber gauge distribution, and an inner surface reinforcement. A variable representing at least one of the tire rubber members of the rubber gauge distribution, the inter-belt rubber gauge distribution, and the belt end rubber gauge distribution, and at least one of the angle, width, cord type, and driving density of each belt layer;
It is characterized by including at least one of a variable representing the structure of one belt portion and a variable representing the shape of at least one pattern of the shape of the block and the position, number, and length of the sipes.

In the present invention, first, in order to predict the performance of a tire design plan (change in tire shape, structure, material, pattern, etc.), the tire design plan is converted into a numerical analysis model. That is, a tire model (numerical analysis model) capable of numerical analysis is created. Then, the road surface related to the target performance is modeled, and the road surface model (numerical analysis model)
Is created, initial conditions are given, a numerical analysis is performed in consideration of the tire and the road surface at the same time, and the target performance is numerically predicted. Whether or not a tire design plan can be determined can be determined from the prediction result. If the result is favorable, the design plan can be adopted, or the performance evaluation of the design plan can be further improved. According to these procedures, the number of times of manufacturing and evaluating the performance of the tire is extremely reduced, so that the efficiency of tire development can be increased.

Therefore, in order to develop a tire based on performance prediction, a numerical analysis model for efficient and accurate tire performance prediction is indispensable. Therefore, in order to predict tire performance, in step (a), a tire model including at least a tire cross-sectional shape including an internal structure and capable of giving a deformation by ground contact, and a road surface model in contact with the tire model Determine. Further, in order to facilitate the analysis of the tire rolling state, an initial condition relating to the rolling of at least one of the tire model and the road surface model is determined. In step (a), an objective function representing a physical quantity for tire performance evaluation, a design variable for determining a tire cross-sectional shape or a tire structure, and a constraint condition for restricting the tire cross-sectional shape or the tire structure are determined.

In the tire model, in addition to the line representing the outer shape of the tire, a line representing the tire crown shape, a belt line representing the belt inside the tire, a carcass line representing the carcass of the tire, and a folding line of the carcass ply inside the tire are provided. Folded ply line to represent, reinforcing material line to represent the line of various reinforcements, gauge distribution of tire rubber members and angle, width of each belt layer to represent the structure of the belt portion,
Block type, block groove wall angle, sipe position, representing code type, implantation density, and pattern shape,
The number and length can be included. The tire model may use a method called a finite element method of dividing the tire model into a plurality of elements, or may use an analytical method. The initial condition is a condition related to the rolling of at least one of the tire model and the road surface model, and includes an initial rotational angular velocity given to the tire model and a moving speed of the road surface model. This initial condition may be any condition as long as a relatively rolling state of the tire can be obtained. Accordingly, the initial condition may be one of the rotation of the tire model or the movement of the road surface model, and may be both the rotation of the tire model and the movement of the road surface model. Also,
The tire model can be provided with an internal pressure such as by air filling, and can be provided with at least one of a rotational displacement and a linear displacement (the displacement may be a force or a speed) and a predetermined load. When friction with the road surface is taken into account, only one of rotational displacement (or force or speed) or linear displacement (or force or speed) may be used.

Further, the tire model may have a pattern partially. By using a tire model having a partial pattern, the calculation time can be reduced as compared with a tire model having a pattern over the entire circumference. Further, the road surface model can reproduce an actual road surface condition by selecting an appropriate value of the friction coefficient μ by DRY, WET, on ice, on snow, on a non-paved surface or the like according to the road surface condition.

As the objective function representing the physical quantity for evaluating the tire performance, use is made of a physical quantity that governs the tire performance such as the tire circumferential belt tension and the lateral spring constant at the time of air filling for improving the steering stability. Can be. Further, as the objective function, a physical quantity that governs the tire performance when the tire is tilted or rotated, such as the linearity of camber thrust, can be used. As a design variable for determining the tire cross-sectional shape, a function representing at least one of a carcass line, a folded ply line, a line representing a tire outer shape, a line representing a tire crown shape, and a reinforcing material line representing various reinforcing material lines. Etc., and the design variables for determining the tire structure include a bead filler gauge distribution, a rubber chafer gauge distribution, a side rubber gauge distribution,
A variable representing the gauge distribution of at least one tire rubber member of the gauge distribution of the tread rubber, the gauge distribution of the tread base rubber, the gauge distribution of the inner surface reinforcing rubber, the gauge distribution of the rubber between belts, and the gauge distribution of the belt end rubber; Variables representing the structure of the belt, bead, and side such as the angle, the width of the belt layer, the height of the ply, the amount of ply folding, the angle, width, position, and material of the bead reinforcing material can be used. Further, as other design variables, variables representing the shape of the pattern such as the shape of the block, the angle of the block groove wall, the position of the sipe, the number, and the length can be used.

Restrictions for restricting the tire cross-sectional shape and the tire structure include a constraint on a peripheral value of a carcass line, a constraint on upper and lower primary natural frequencies, a constraint on an angle of a belt layer, a width of a belt layer, a tire dimension, and a spring. There are constants, tire deformation, tire weight, stress, strain, strain energy, and restrictions on rolling resistance. It should be noted that the objective function, the design variables, and the constraints are not limited to the above examples, and various ones can be determined according to the tire design purpose.

Next, in step (b), tire performance is predicted based on physical quantities generated in the tire model due to deformation of the tire model. In the prediction in step (b), if the tire model contact area and the contact pressure are physical quantities, the tire W
ET performance can be predicted. Further, if at least one of the contact area, the contact pressure, and the shearing force on at least one of the icy road surface and the snowy road surface of the tire model is used as a physical quantity, it is possible to predict the performance on the tire ice and snow. In the prediction of the step (b), the deformation calculation of the tire model is executed in the step (e), the giving condition is given to the tire model in the step (f), and the deformation calculation in the step (e) is performed in the step (g). To the tire model of step (f), and by the applied condition,
Calculation is performed until the tire model is in a transient state, and a physical quantity generated in the tire model is obtained in step (h), and tire performance can be predicted based on the physical quantity in step (i).

When predicting tire performance, it is necessary to make a prediction in accordance with the possible behavior of the tire. For this reason, when executing the deformation calculation of the tire model, it is preferable to add an application condition. The giving condition is a condition in which the tire is tilted or rotated in a state in accordance with a behavior that the tire can take, and there is a condition in which a camber angle and a thrust angle are given to the rotation axis of the tire model. By applying the conditions for the deformation to the calculation of the deformation of the tire model, the tire model can be calculated in a state in accordance with the behavior that the actual tire can take. This makes it easy to predict tire performance such as steering stability and camber thrust linearity.

In the deformation calculation of the tire model in the step (e), a deformation calculation when a deformation is given by the ground contact can be executed. Further, in the deformation calculation of the tire model, iterative calculation can be performed. In the deformation calculation of the tire model, the elapsed time of the predetermined time for performing the repeated calculation can be 10 msec or less, preferably 1 msec or less, more preferably 1 μ · sec or less. If the elapsed time is too long, the pseudo-flow state does not match the behavior of the tire in contact with the road surface model, and the accuracy of the numerical model deteriorates. For this reason, it is necessary to adopt an appropriate value for the elapsed time.

Next, in step (c), the value of the design variable that gives the optimum value of the objective function is determined in consideration of the predicted tire performance and constraints. In this case, the constraint is based on the sensitivity of the objective function, which is the ratio of the change amount of the objective function to the unit change amount of the design variable, and the sensitivity of the constraint condition, which is the ratio of the change amount of the constraint condition to the unit change amount of the design variable. The amount of change of the design variable that gives the optimum value of the objective function is predicted while considering the conditions, and the value of the objective function and the design variable when the design variable is changed by the amount corresponding to the predicted amount are the amounts corresponding to the predicted amount. It is effective to calculate the value of the constraint condition at the time of the change, and obtain the value of the design variable that gives the optimal value of the objective function while considering the constraint condition based on the predicted value and the calculated value. As a result, the value of the design variable when the value of the objective function is optimized in consideration of the predicted tire performance and the constraint condition is obtained.

As a result of various studies, the present inventors have focused on applying the “genetic algorithm method” used in a different field to a special field called a tire.
Tried all kinds of studies, predicting tire performance in the course of the genetic algorithm, that is, predicting tire performance such as steering stability and linearity of camber thrust, enabling transient analysis especially for tires that tilt when tires touch down,
In order to streamline tire development and facilitate the provision of tires with good performance, it was specifically established as a tire design method.

More specifically, the invention of claim 11 is based on claim 1
In the pneumatic tire designing method of (1), in the step (a), a selection target group including a plurality of tire models including at least a tire cross-sectional shape including an internal structure and capable of being deformed by grounding is defined. For each tire model of the selection target group, an objective function representing a tire performance evaluation physical quantity, a design variable for determining a tire cross-sectional shape or a tire structure or a pattern shape, a tire cross-sectional shape, a tire structure, a pattern shape, and a performance evaluation physical quantity. And a constraint that restricts at least one of the tire dimensions, and an objective function and an adaptive function that can be evaluated from the constraint, and a road surface model that contacts the tire model, and at least one of the tire model and the road surface model Initial conditions related to the rolling of the model are determined. Selecting two tire models from the selection target group based on the above, generating a new tire model by crossing design variables of each tire model with a predetermined probability, and a part of design variables of at least one tire model. And at least one of generating a new tire model by changing the tire model, obtaining the objective function, constraint, and adaptation function of the tire model in which the design variables are changed, and calculating the tire model and the tire in which the design variables are not changed. The model is saved and repeated until the number of saved tire models reaches a predetermined number, and it is determined whether a new group of the stored predetermined number of tire models satisfies a predetermined convergence condition. A group is used as the selection target group, and the selection target group is repeated until the predetermined convergence condition is satisfied. Determine the value of the design variable which gives the optimum value of the objective function while considering the constraint condition among the predetermined number of tire models stored when plus, characterized in that.

According to a twelfth aspect of the present invention, there is provided the pneumatic tire designing method according to the eleventh aspect, wherein in the step (c), the tire model in which the design variable has been changed is determined based on a unit change amount of the design variable. The optimum value of the objective function is given based on the sensitivity of the objective function, which is the ratio of the change amount of the objective function, and the sensitivity of the constraint condition, which is the ratio of the change amount of the constraint condition to the unit change amount of the design variable, while considering the constraints. In addition to predicting the amount of change in the design variable, the value of the objective function when the design variable is changed by the amount corresponding to the predicted amount and the value of the constraint condition when the design variable is changed by the amount corresponding to the predicted amount are calculated. An adaptive function is obtained from the value of the objective function and the value of the constraint to store the tire model and the tire model in which the design variables are not changed, and the number of stored tire basic models becomes a predetermined number. And repeating until.

In the present invention, first, in the step (a), a selection target group including a plurality of tire models having at least a tire cross-sectional shape including an internal structure and having a pattern shape capable of being deformed by grounding. For each tire model of the selection target group, an objective function representing a tire performance evaluation physical quantity, a design variable for determining a tire cross-sectional shape or a tire structure, a tire cross-sectional shape, a tire structure, a performance evaluation physical quantity, and a tire dimension And a road surface model that is in contact with the tire model and a rolling condition of at least one of the tire model and the road surface model is determined. And the initial conditions related to In the step (c), a new tire model is generated by selecting two tire models from the selection target group based on the adaptation function and crossing design variables of the respective tire models at a predetermined probability. At least one of generating a new tire model by changing a part of the design variables of at least one tire model is performed, and an objective function, a constraint condition, and an adaptive function of the tire model in which the design variables are changed are obtained. The tire models and the tire models in which the design variables are not changed are stored and repeated until the stored number of the tire models reaches a predetermined number. Judgment is made, and when the convergence condition is not satisfied, the new group is set as the selection target group, and the new selection group remains in the predetermined convergence condition. With repeated, it is also effective for obtaining the value of the design variable which gives the optimum value of the objective function while considering the constraint condition among the predetermined number of tire models stored when filled with the predetermined convergence condition.

In this case, in step (c), for the tire model in which the design variables are changed, the sensitivity of the objective function, which is the ratio of the change of the objective function to the unit change of the design variables, and the change in the unit change of the design variables. Based on the sensitivity of the constraint, which is the ratio of the amount of change in the constraint, the change in the design variable that gives the optimal value of the objective function is predicted while considering the constraint, and the design variable is changed by an amount corresponding to the predicted amount. The value of the objective function and the value of the constraint when the design variable is changed by an amount corresponding to the predicted amount are calculated, and the adaptive function is obtained from the value of the objective function and the value of the constraint to obtain the tire model and the design. It is more effective to store the tire basic model in which the variables are not changed and repeat until the stored tire basic model reaches a predetermined number. Also in this case, the value of the design variable when the value of the objective function is optimized in consideration of the constraint condition is obtained. In addition, as the adaptation function that can be evaluated from the objective function and the constraint condition, a function that determines the fitness for the tire model from the objective function and the constraint condition can be used. Also, the objective function, design variables, constraints and adaptation functions are not limited to the above examples,
Various things can be determined according to the tire design purpose. Further, as for the crossover of the design variables of the tire basic models, there is a method of exchanging design variables of a part or a predetermined portion or later of the design variables of the two selected tire models. Further, as a method of partially changing the design variables of the tire model, there is a method of changing (mutating) the design variables at positions determined by a predetermined probability or the like.

Further, the present inventor has made various studies, and as a result, has found that "non-linear prediction technology in which a neural network of a higher animal is modeled by engineering, for example, a neural network" and "optimum" which are used in different fields. Focusing on the application of the "customized design method" to the special field of tire design, we tried to study and established a concrete tire design method.

More specifically, as described in the thirteenth aspect of the present invention, the method of designing a pneumatic tire according to the first aspect, wherein in the step (a), the tire cross-sectional shape including the internal structure is changed. Define a conversion system that associates at least the non-linear correspondence between the design parameters of the tire model capable of giving deformation by ground contact and the performance of the tire,
A constraint condition for limiting at least one of the allowable range of the tire performance and the tire manufacturing condition is defined as the constraint condition, and in the step (c), the objective function is determined by using the conversion system defined in the step (a). And determining a tire design parameter that gives an optimal value of the objective function based on the constraint condition. In the step (d), the tire is designed based on the tire design parameter.

According to a fourteenth aspect of the present invention, there is provided the pneumatic tire designing method according to the thirteenth aspect, wherein in the step (c), the design parameters of the tire model are determined as design variables, and the constraint conditions are considered. The value of the design variable that gives the optimal value of the objective function is determined using the conversion system determined in step (a), and the tire is designed based on the design variable that gives the optimal value of the objective function in step (d). It is characterized by the following.

The invention of claim 15 is the invention of claim 13 or 1
5. The method for designing a pneumatic tire according to claim 4, wherein the conversion is performed using data of a multilayer feedforward neural network learned to convert the design parameters of the tire model into the tire performance in the step (a). It is characterized by comprising a system.

The values of tire performance, for example, steering stability, are determined by tire design parameters, for example, the tire cross-sectional shape including the internal structure and the tire structure. However, in many cases, the tire performance does not change linearly even when the values of the tire cross-sectional shape and the tire structure are changed linearly. In view of this, in step (a) of the present invention, a conversion system is defined which relates at least a non-linear correspondence between the design parameters of a tire model including at least a tire cross-sectional shape including an internal structure and capable of giving deformation by grounding and tire performance. ing. In addition, a constraint condition for limiting at least one of the allowable range of the tire performance and the tire manufacturing condition is defined as the constraint condition. This conversion system can be determined by using a non-linear prediction technique in which a neural network such as a neural network is engineered.

In the step (c), using the conversion system determined in the step (a), a design parameter of a tire model that gives an optimum value of the objective function is obtained based on the objective function and the constraint condition. The tire is designed based on the design parameters of the tire model. As a result, a conversion system for associating the non-linear correspondence between the tire design parameters and the tire performance is determined, and the conversion system can find a mutual relationship in which the correspondence between the plurality of tire design parameters and the performance is associated. . Therefore, by obtaining tire design parameters that give the optimum value of the objective function and designing the tire based on the design parameters, a high-performance tire can be designed.

When designing a tire, the design parameters of the tire (tire model) are determined as design variables, and the optimum value of the objective function is given by using the conversion system determined in step (a) while considering the constraints. , And the tire can be designed based on the design variables that give the optimal value of the objective function. As described above, by taking into account the constraint conditions, it is possible to consider the allowable range of at least one of the tire performance and the tire design parameter, and to specify the design range in advance or set a desired range.

When the value of the design variable is obtained, the sensitivity of the objective function, which is the ratio of the change amount of the objective function to the unit change amount of the design variable, and the ratio of the change amount of the constraint condition to the unit change amount of the design variable. The amount of change of the design variable that gives the optimal value of the objective function while considering the constraint condition based on the sensitivity of the constraint condition, and the value of the objective function when the design variable is changed by an amount corresponding to the predicted amount And calculate the value of the constraint condition when the design variable is changed by an amount corresponding to the predicted amount, and use the conversion system determined in step (a) based on the predicted value and the calculated value while considering the constraint condition. It is effective to determine the value of the design variable that gives the optimum value of the objective function. Thereby, the value of the design variable when the value of the objective function is optimized in consideration of the constraint condition is obtained. Then, the tire can be designed by changing the tire design parameters and the like based on the design variables giving the optimal value of the objective function.

Further, according to the present invention, a tire is designed such that the block height in each block existing in the tread pattern is uniquely optimized in accordance with the contact pressure with respect to the input received by the tire. Alternatively, the unevenness of the ground pressure in the block can be eliminated.
That is, in the case of the tread shape in which the block height is set to be constant, the contact pressure distribution characteristics become non-uniform in the block. According to the present invention, by changing the shape of the tread, it is possible to obtain a contact pressure distribution characteristic that is substantially uniform in the block.

For this purpose, first, a basic shape model is created and input conditions are given to it. The objective function value is calculated from the contact pressure obtained at this time. Then, the shape is changed using the design variable as the tread surface shape of the block, and an input is again provided to the model using the new shape to obtain the distribution of the contact pressure. This operation is repeated until the optimal value of the objective function is given. Note that the present invention can be applied to a case where there is a constraint condition. In this case, an objective function value and a constraint function value are obtained after an input is given. The end of the design in this case is when the optimum value of the objective function is obtained within the constraint conditions. The tire is designed based on the optimal design variables.

More specifically, a pneumatic tire designing method according to a sixteenth aspect of the present invention is the pneumatic tire designing method according to the first aspect, wherein in the step (a),
One of the shapes of the block including the internal structure, the pattern shape of one portion of the tire crown portion including the internal structure, and the selected shape of the land portion continuing in the tire circumferential direction including the internal structure is selected. Further defining a shape basic model to represent, providing at least one input condition to the shape basic model, and a tread shape representing at least a part of the shape of a single unit of the block or the shape of a pattern or the shape of a land portion as a design variable, The tire contact pressure under the input condition is calculated and determined as an objective function.

According to a seventeenth aspect, in the pneumatic tire designing method according to the sixteenth aspect, at least one of a tire contact area and a change range of a design variable is further defined as the constraint condition, The step (c) is characterized in that the value of the design variable is changed until the optimum value of the objective function is given while considering the constraint condition.

The invention of claim 18 is the method of designing a pneumatic tire according to claim 17, wherein the design variables are:
It is characterized in that the design variables of at least one of a portion higher and a portion lower than the tire average contact pressure are changed.

In the present invention, the shape of the block including the internal structure alone, the pattern shape of one portion of the tire crown portion including the internal structure, and the shape of the land portion extending in the tire circumferential direction including the internal structure are included. A shape basic model representing one selected shape is determined.

The basic shape model representing the shape of a single block can be composed of a function representing a line for specifying the outer shape of the single block or a variable representing the coordinate value of an inflection point. In addition, as a shape basic model representing a pattern shape of a part of the tire crown portion including the internal structure, a pattern shape on a road surface ground side of one land portion of the tire crown portion can be geometrically analyzed. It can be composed of functions, for example, functions for defining polygons such as rectangles and rhombuses. In addition, the shape basic model representing the shape of the land portion continuous in the tire circumferential direction including the internal structure can be constituted by a function representing a line representing a tire sectional shape or a variable representing a coordinate value of an inflection point. These shape basic models are
A model based on a method called a finite element method that divides into a plurality of elements may be used, or a model based on an analytical method may be used.

In addition, at least one of the shape basic models
Give two input conditions. The input conditions include a load condition indicating a load to be added and a direction condition indicating a shear direction. next,
The tread shape representing at least a part of the shape of the block alone or the shape of the pattern or the shape of the land portion is used as a design variable, and the tire contact pressure under input conditions is calculated and determined as an objective function.

In step (c), the value of the design variable to which the optimal value of the objective function is given can be obtained, and the calculation is performed while changing the value of the design variable until the optimal value of the objective function is given. Thus, the value of the design variable can be obtained.

When the tread shape is used as a design variable and the tire contact pressure under input conditions is calculated and determined as an objective function, it is preferable to consider constraints to reduce the calculation load. Therefore, at least one of the tire contact area and the change range of the design variable is further defined as the constraint condition, and in step (c), the value of the design variable is changed while giving the optimum value of the objective function while considering the constraint condition. The change range of the design variable can be represented by any of the range of the tread and the block height.

When the value of the design variable to which the optimum value of the objective function is given is obtained, the design variable at least one of a portion higher and lower than the tire average contact pressure can be changed. As described above, when a design variable at a location higher than the tire average contact pressure is changed, the change can be made so as to reduce the block height. When the design variable at a location lower than the tire average contact pressure is changed, it can be changed so as to increase the block height. Further, the design variables can be changed by changing the block height in accordance with each deviation from the tire average contact pressure.

When there are a plurality of input conditions,
An input condition with a large change amount of the block height can be preferentially changed.

Further, at least a part of the tread shape represented by the design variable can be represented by any one of a polynomial, a piecewise polynomial, a spline function and a rational function. In the case where at least a part of the tread shape represented by these formulas is changed, it may be performed for each calculation or once for each calculation in consideration of the calculation time and the capability of the computer.

Incidentally, by forming the tire by a structure or the like based on the design parameters of the tire designed by the above-described tire design method, the formed tire is constituted by the design parameters having the best performance. The contents of the optimal design parameters can be directly determined by the application conditions such as the manufacturing conditions and the cost.

The above-described tire designing method includes storage means for storing a program for executing the processing of steps (a) and (b) according to claims 1 to 18, and the program is stored in the storage means. Read and
Prediction means for predicting the performance of the tire by physical quantities generated in the tire by deformation using the design parameters of the tire, conversion system calculation means for determining a non-linear correspondence between the tire design parameters and the performance of the tire, An input function for defining an objective function representing tire performance, defining a constraint condition for restricting at least one of the allowable range of the tire performance and the tire manufacturing condition, and inputting as an optimization item, and the conversion system calculating means. And an optimization calculation device for obtaining a tire design parameter that gives an optimal value of the objective function based on the optimization item input by the input device.
According to a twentieth aspect of the present invention, in the optimization analysis apparatus according to the nineteenth aspect, the conversion system calculating means has been trained to convert the design parameters of the tire into the tire performance. It is characterized by being.

The conversion system calculating means can obtain a nonlinear correspondence between the tire design parameters and the tire application conditions and the tire performance. These application conditions include manufacturing conditions for forming the tire, tire weight, or overall cost. Further, the conversion system calculating means may be configured by a multilayer feedforward neural network that has been learned to convert the tire design parameters into the tire performance.

The above-described tire designing method can provide a storage medium storing an optimization analysis program which can be easily carried by a storage medium including a program according to the following procedure. That is, a storage medium storing a tire optimization analysis program for designing a tire by a computer, wherein the optimization analysis program performs the processing of steps (a) and (b) of claims 1 to 18. Including a prediction program for executing the, read the prediction program, and predict the performance of the tire by the physical quantity generated in the tire by deformation using the tire design parameters, the relationship between the tire design parameters and the performance of the tire A non-linear correspondence is defined, an objective function representing the performance of the tire is defined, and a constraint condition that restricts an allowable range of at least one of the performance of the tire and a manufacturing condition of the tire is determined. Tire design parameters that provide an optimal value of the objective function based on the objective function and the constraint conditions The seeking to design a tire based on the design parameters of the tire.

[0062]

Embodiments of the present invention will be described below in detail with reference to the drawings. In this embodiment, the present invention is applied to the design of a pneumatic tire.

FIG. 1 schematically shows a personal computer for carrying out the pneumatic tire designing method of the present invention. This personal computer calculates a design variable that satisfies constraints and optimizes an objective function, for example, a maximum or a minimum, while predicting the performance of a tire according to a keyboard 10 for inputting data and the like according to a processing program stored in advance. Computer main body 12 and CRT 1 for displaying calculation results and the like of computer main body 12
4.

The computer main body 12 has a floppy disk unit (FDU) into which a floppy disk (FD) as a recording medium can be inserted and removed. The processing routine and the like described later can be read from and written to the floppy disk FD using the FDU. Therefore, a processing routine described later may be recorded in the FD in advance, and the processing program recorded in the FD may be executed via the FDU. Also, a large-capacity storage device (not shown) such as a hard disk device is connected to the computer main body 12, and the processing program recorded in the FD is stored (installed) in the large-capacity storage device (not shown) and executed. Is also good. The recording medium is a CD-RO
There are disk recording media such as M, DVD, MD, MO, etc.
When these are used, a CD-ROM device, a DVD device, an MD device, an MO device, or the like may be used instead of the FDU or further.

[First Embodiment] Next, in order to improve the steering stability, the first embodiment of designing the outer shape of the tire to optimize the linearity of the camber thrust while predicting the performance of the tire. The form will be described.

FIG. 2 shows a processing routine of the program of the first embodiment. In step 100, an initial model such as a tire model for converting a design plan of a tire to be designed (change of a tire shape, a structure, a material, a pattern, and the like) into a model for numerical analysis, and a road surface model for tire performance evaluation. The creation process is executed.

In step 100, the processing routine shown in FIG. 3 is executed. In step 130 of FIG. 3, a design plan of the tire to be evaluated (change of tire shape, structure, material, pattern, etc.) is determined. Then, in step 136, the road surface state is input together with the preparation of the road surface model, and in the next step 138, the initial conditions given to the tires are set.

More specifically, in step 130 of FIG. 3, an initial tire design plan (change of tire shape, structure, material, pattern, etc.) is determined. Create a tire model to put it into the model.

The creation of the tire model differs slightly depending on the numerical analysis method used. In this embodiment, a finite element method (FEM) is used as a numerical analysis method. Therefore, the tire model created in step 132 is
Element division corresponding to the finite element method (FEM), for example, divided into a plurality of elements by mesh division, and the tires are converted into numerical data in the form of input data to a computer program created based on numerical and analytical methods Say. This element division refers to dividing an object such as a tire and a road surface into several small (finite) small parts. After performing the calculation for each of the small parts and calculating for all the small parts, the total response can be obtained by adding all the small parts. Note that a difference method or a finite volume method may be used as the numerical analysis method.

In the creation of a tire model in step 132, a pattern is modeled after a model of a tire cross section is created. Specifically, a tire model creation routine shown in FIG. 4 is executed. First, in step 140, a model of a cross section in the tire radial direction is created. That is, tire section data is created. This tire cross section data
The outer shape of the tire can be measured with a laser shape measuring instrument or the like and the value can be collected. In addition, an accurate internal structure of the tire is obtained from a design drawing, actual tire sectional data, and the like. Rubber in the tire section, teaching materials (belts, plies, etc.
A reinforcing cord made of iron, organic fiber, or the like is bundled in a sheet shape), and each is modeled according to the modeling method of the finite element method. FIG. 5 shows a model of the tire radial cross section modeled in this manner. In the next step 142, 2
The three-dimensional (3D) model of the tire is created by developing the tire cross-section data (model of the cross section in the tire radial direction), which is the dimensional data, for one round in the circumferential direction. In this case, it is desirable that the rubber part is modeled by an eight-node solid element and the teaching aid is modeled by an anisotropic shell element capable of expressing an angle. For example, the rubber part can be handled by an 8-node solid element as shown in FIG. 6 (A), and the reinforcement (belt, ply) is handled as a shell element as shown in FIG. 6 (B). The angle θ of the reinforcing material can be considered two-dimensionally. Next Step 1
At 44, the pattern is modeled. This pattern is modeled by any of the following procedures.

Procedure: A part or the whole of the pattern is separately modeled and attached to the tire model as a tread portion. Procedure: A pattern is created in consideration of the rib / lag components when developing the tire cross-section data in the circumferential direction.

FIG. 7 shows a pattern modeled by the above procedure. In the drawings, the rotation axis 31 of the tire model is shown. The rotating shaft 31 is used when applying conditions to the tire model, as described later, and by applying a force to the rotating shaft 31, the camber angle CA and the slip angle SA can be generated. it can. The slip angle SA is an angle formed by the tire model with respect to the traveling direction of the tire model (vehicle), as shown in FIG. 8A, and is a condition for rotating the rotating shaft 31 in a plane along a road surface. It can be caused by giving. The camber angle CA is shown in FIG.
As shown in FIG. 7, the angle is an angle formed when the tire model is tilted with respect to the vertical direction of the road surface, and can be generated by giving a condition for rotating the rotary shaft 31 in a plane perpendicular to the road surface.

As an example of the preparation of the tire model in step 132, the tire cross-sectional shape in a natural equilibrium state is used as a reference shape, and this reference shape is used as a numerical value to calculate the tire circumferential belt tension at the time of air filling as in the finite element method. The model is modeled by a method that can be obtained analytically and analytically, a tire cross-sectional shape including an internal structure is represented, and a tire basic model divided into a plurality of elements by mesh division is obtained.
The reference shape is not limited to the tire cross-sectional shape in a natural equilibrium state, and may be any shape. Here, the model creation means to digitize the tire shape, structure, material, and pattern in a form of input data to a computer program created based on a numerical / analytical method.

FIG. 9 shows this basic tire model, where CL is a carcass line, OL is a line (tread line) representing the outer shape of the tire (tread outline), PL is a ply line, and B1 and B2 are lines representing a belt. Are respectively shown. The tire basic model is divided into a plurality of elements by a plurality of normals NL1 _, NL2 _, NL3 _, ... Of the tread line OL. In the above description, an example in which the tire basic model is divided into a plurality of elements by a plurality of normals of a tread line has been described. It may be divided, or may be divided into an arbitrary shape such as a triangle depending on the design purpose.

After the tire model is created as described above, the process proceeds to step 136 in FIG. 3, where the road surface state is input together with the creation of the road surface model. This step 10
Reference numeral 6 denotes an input for modeling a road surface and setting the modeled road surface to an actual road surface state. In modeling the road surface, the road surface shape is modeled by dividing the road surface shape into elements, and the road surface state is input by selectively setting the friction coefficient μ of the road surface. That is, since there is a friction coefficient μ of the road surface corresponding to dry (DRY), wet (WET), on ice, snow, unpaved, etc. depending on the road surface condition, by selecting an appropriate value for the friction coefficient μ, The road condition can be reproduced. Also, the road surface model only needs to be in contact with at least a part of the tire model, and its shape need not be a planar state. As an example of a road surface model, a model of a road surface such as a chassis dynamo is shown in FIG. The tire model shown in FIG. 7 is grounded to the road surface model shown in FIG.

When the road surface condition is input in this way, initial conditions are set in step 138 of FIG. To analyze the tire model, it is preferable to simulate the behavior of the tire by giving the tire model initial conditions for analysis in order to reduce the calculation load.
Therefore, in the present embodiment, the initial angular velocity is given to the rolling state, that is, the tire model, instead of considering the tire model in the stopped state.

In setting the initial conditions in step 138 of FIG. 3, the processing routine of FIG. 15B is executed.
First, the routine proceeds to step 150, where an initial angular velocity is given to the tire model. After the initial angular velocity is given to the tire model, in a next step 152, an internal pressure is applied to the tire model, and in a next step 154, at least one of a rotational displacement and a rectilinear displacement (the displacement may be force or speed) is applied to the tire model. And a predetermined load. When friction with the road surface is taken into account, only one of rotational displacement (or force or speed) or linear displacement (or force or speed) may be used.

When the initial model creation processing (step 100) is completed as described above, the process proceeds to step 102 in FIG. 2, where the objective function representing the physical quantity for evaluating the tire performance, the constraint conditions for restricting the tire sectional shape, and Determine design variables that determine the tire cross-sectional shape.

In the present embodiment, in order to improve the steering stability, the shape of the tread line that optimizes the linearity of the camber thrust is designed. Has been determined. Note that the constraint condition G specifies that the height Th and width Tw (see FIG. 12A) of the shape of the tread line are within +5 mm. In addition, it is possible to set only a numerical limit necessary for calculation without setting a special constraint condition. In the case of optimizing the carcass line, the constraint condition G may be set to “the peripheral value of the carcass line is within ± 5% of the peripheral value of the carcass line of the basic tire model”. In this case, the peripheral value of the carcass line can be calculated as the sum of the distances between the nodes (intersections of the carcass line and the normal) of the carcass line within the range in which the tire shape is changed.

As the design variables for determining the tire cross-sectional shape, the shape of the tread line OL is adopted as shown in FIG. The shape of the tread line can be determined by the approximation routine of FIG. In step 160 of this approximation routine, the tire outer shape in the tire cross-sectional shape is determined. That is, the equator length (diameter) of the tire model is calculated. In the next step 162, the intersection of the tread line and the straight line whose length is determined on the outer shape of the tire is set as the reference point P. As shown in FIG. 12B, a reference point P is set at a position at a distance of radius R1 from the center point O of the tire model. In the next step 164, a predetermined distance s (for example, 0.1 m) from the set reference point P
m) Moved in the direction along the tread line to set the contact point Q (s), and in the next step 166, the distance R from the tread line according to the following equation (1).
(S) is calculated. The straight line along the distance R (s) matches the normal at each contact point Q (s). Next Step 1
At 68, the obtained distance R (s) is set to the design variables r ₁ ,
r ₂ , r ₃ ,... (hereinafter represented as r _{i in the} general formula.
i is equivalent to the number of contacts, and i = 1, 2,..., the maximum value). The contact Q (s) is also
Expressed as Q _{i in the} general formula. R (s) = a · s 2 + b · s + c ··· (1) where, a, b, c are constants

The cross section of the tire is tread.
There is a side part other than the line, and the shape of this side part is also set.
It may be added to the total variable. In detail, the side part,
According to the Lagrange interpolation routine of FIG.
Can be determined. Lagrange interpolation routine
In step 170 of FIG.
A reference point P is set inside the frame. In the next step 172
Is the node q near the belt end₁Near the rim
Node q of_TwoThe range up to the range that changes the tire shape
To specify. In step 174, the node q₁And reference point
A straight line connecting P and a reference line is referred to as a reference line.
_TwoAngle θ, which is an angle formed by a straight line connecting
Is calculated, and in step 176, the approximate expression (dθ = θ /
The order of Lagrange interpolation: The order of Lagrange interpolation
(The number must be input by the user in advance.)
Calculate the minute dθ. In step 178, FIG.
As shown in the figure, a temporary
Imaginary line L_1,L_2,L_3,... and the method closest to the virtual line
Line nl_1,nl_2,nl_3,... is selected. Next steps
At 180, as shown in FIG.
Line nl_1,nl_2,nl_3,... Top innermost node Q_{j + 1,}Q
_{j + 2,}Q_{j + 3,}... and the distance r between the reference point P_{j + 1,}r _{j + 2,}r
_{j + 3,}... (however, j is the maximum value of i and the final value of j
The value is the order of Lagrange interpolation -1). And
In step 182, the above distance is used as a design variable.
Set to.

[0082] After determining the objective function OBJ in this manner, constraint condition G and the design variables Rn (or even adding r _i), at step 104 of FIG. 2, an initial tire performance prediction processing is executed. In this tire performance prediction process, the process routine of FIG. 16 is executed, and based on the numerical models created or set so far, the analysis of the deformation calculation of the tire model described in detail below is analyzed. Do. In this embodiment, in order to obtain a transient state,
Deformation calculation of the tire model is performed within an elapsed time of 1 msec or less, and an application condition is provided.

In step 200 of FIG. 16, a deformation calculation (analysis) of the tire model is performed, and in the next step 202, a condition is given to the tire rotation axis while the rotating surface of the tire model is maintained. In the next step 204, it is determined whether or not the elapsed time is within 1 msec.
Returning to 0, the deformation calculation is performed again. If the result is negative in step 204, the process proceeds to step 206, where it is determined whether the calculation is completed. If negative in step 206, the process returns to step 200, and if positive in step 206, the process proceeds to step 208.

The deformation calculation of the tire model is a calculation of the deformation of the tire model based on the finite element method, and the tire model is given a condition to be applied to the tire rotation axis while maintaining the rotating surface. And to get the transitional state,
While the elapsed time (single elapsed time) is 1 msec or less, the deformation calculation of the tire model is repeated. By repeating this every 1 msec, it is possible to simulate a transient state related to tire performance prediction. Here, 1 msec is a time capable of sufficiently expressing the process in which the pattern in the ground contact surface is deformed by rolling of the tire. The application condition is to apply the condition to the tire rotation axis while maintaining the rotation surface, and in the present embodiment, to change the position of the rotation axis (not shown) of the tire model in order to generate the slip angle and the camber angle. (Details will be described later).

In the above calculation (analysis), a case has been described in which the repeated calculation is performed during a preferable elapsed time in which the elapsed time (single elapsed time) is 1 msec or less. In the present invention, the elapsed time is reduced to 1 msec. There is no limitation, and an elapsed time of 10 msec or less can be employed, preferably 1 msec or less, and more preferably 1 μsec or less. In this case, the repetition time (single elapsed time) to which the provision condition is provided is determined. However, the present invention is not limited to the provision of the provision condition, and may be used only for the deformation calculation.

The following example is used to determine whether or not the calculation is completed in the next step 206. If the tire model is a rolling model with an all-around pattern, the calculation is repeatedly performed until the target physical quantity (such as camber thrust) can be regarded as a steady state (a state where the physical quantity can be regarded as the same as the previously calculated physical quantity), and the calculation is completed. If so, a positive determination is made. Furthermore, it is also possible to end the processing when a predetermined time has elapsed.
The predetermined time in this case is preferably 100 msec or more,
More preferably, it is 300 msec or more.

If the tire model models only a part of the pattern, the calculation is repeated until the deformation of the pattern part to be analyzed is completed. The deformation of the pattern portion refers to the deformation until the pattern portion comes into contact with the road surface model after rolling and then separates from the road surface model, or the rolling portion causes the pattern portion to come into contact with the road model after contacting the fluid model. . The deformation of the pattern portion may be performed when the tire comes into contact with each of the models after rolling one or more revolutions. Furthermore, it is also possible to end the processing when a predetermined time has elapsed. The predetermined time in this case is preferably 100 msec or more, more preferably 300 msec.
That is all.

Next, the details of step 202 in FIG. 16 will be described. In the processing of adding the application condition to the tire model in step 202, the processing routine of FIG. 17 is executed. First, in step 220, it is determined whether the slip angle SA to be changed is determined by determining whether the slip angle SA to be changed is set.
If the result in step 220 is negative, step 2
Proceeding to 24, if affirmative, step 222 changes the slip angle SA. In this step 222, the rotation axis 31 of the tire model is rotated by a predetermined set value in a plane along the road surface with respect to the traveling direction of the tire model (vehicle) (see FIG. 8A). In the next step 224, it is determined whether or not the camber angle CA to be changed has been set.
It is determined whether or not A is changed. If the result of the determination in step 224 is negative, the present routine is terminated. If the result is affirmative, the camber angle CA is changed in step 226. In this step 226, the rotation axis 31 of the tire model is rotated by a predetermined set value to incline the tire model with respect to the vertical direction of the road surface (FIG. 8).
(B)).

As described above, the slip angle S is added to the tire model.
By giving A and the camber angle CA, the tire model can be calculated in consideration of the running state, and the performance prediction can be performed efficiently.

After the analysis and the provision of the provision condition are performed in this way, the process returns to the analysis, and the calculation is performed under the changed provision condition. This is repeated until the calculation is completed. When the calculation is completed, the result in step 206 in FIG. 16 is affirmative, the process proceeds to step 208, and the calculation result is output (evaluated) as a prediction result.

In the above description, a case has been described in which the analysis and the assignment of the assignment condition are repeated, and the calculation result is output (evaluated) when the calculation is completed. However, during the repeated calculation, the calculation result at that time is output. The output may be evaluated or may be evaluated sequentially. That is, output and evaluation may be performed during calculation.

As the output of the prediction result, values or distributions of camber thrust, contact pressure, contact area, contact shape, energy, and the like can be adopted. As specific examples of the output of the prediction result, output and visualization of camber thrust,
And output and visualization of the distribution related to the contact portion between the tire model and the road surface model. Here, if both the periphery of the tire model and the periphery of the pattern are represented by a diagram or the like, they can be visualized. Further, in visualizing the distribution, the periphery of the tire model and the periphery of the pattern may be created as a diagram, and the values may be displayed on the figure in association with the colors and patterns.

In addition, the evaluation is a subjective evaluation (judgment by a skilled person, etc.), whether the camber thrust, the contact pressure, the contact area, the contact shape, the energy locally rises or fluctuates, or a sufficient value. Is obtained, or the like can be adopted.

The evaluation of the prediction result is based on the output value of the prediction result and the distribution of the output value. The evaluation value can be determined by expressing it in a typical manner.

Therefore, it is possible to judge whether or not the prediction performance is good from the evaluation of the prediction result. This determination may be made by input from a keyboard.Also, an allowable range is predetermined for the evaluation value, and when the evaluation value of the prediction result is within the allowable range, it is determined that the prediction performance is good. You may make it.

As a result of the evaluation of the predicted performance, if the target performance is not sufficient, the processing is stopped at this point, and the design plan is changed (corrected), and then the tire design is started again (the processing performed so far). May be redone), or the result of the evaluation of the prediction performance may be stored and referred to at the time of optimization.

[0097] As described above, when the initial tire performance prediction process ends, the process proceeds to step 106 in FIG. 2, the initial value of the initial value OBJo and constraints G of the objective function OBJ at an initial value ro of the design variables r _i Go is calculated.

[0098] In the next step 108, is changed by each [Delta] r _i design variables r _i in order to change the tire basic model, the next step 110, determines the tire correction model. In step 110, the distance s of the variable r _m between nodes and the reference point P other than the node Q _i corresponding to the design variable
Can be calculated by the following equation. in this case,
θ _m is an estimated distance between the node and the reference point P.
Incidentally, when optimizing the side portion, may be step 110, the distance r _m between the innermost node and the reference point P other than the node Q _i corresponding to the design variable is calculated by the following equation. In this case, θ _m is an estimated angle of a straight line connecting the node and the reference point P from the reference line.

[0099]

(Equation 1)

In step 110, nodes other than the other nodes on the normal (nodes on the tread line OL indicating the outer shape of the tire), that is, the lines B1 and B2 indicating the carcass line, the ply line PL, and the belt obtains distances between the node Q _i of the nodes, seek other node coordinates on normal by adding the distance obtained in the coordinates of the node Q _i on the tread line OL, were [Delta] r _i changed design variables A later tire cross-sectional shape, that is, a tire correction model is determined.

After the tire correction model is determined in this way, in the next step 112, the above-described step 1 is performed.
In the same manner as in step 04, the tire performance prediction processing of the corrected tire model is executed. When the tire performance prediction processing is executed in step 112, it is possible to determine whether the prediction performance is good or not from the evaluation of the prediction result in the same manner as described above. This determination may be made by input from a keyboard.Also, an allowable range is predetermined for the evaluation value, and when the evaluation value of the prediction result is within the allowable range, it is determined that the prediction performance is good. You may make it. Also, as a result of the evaluation of the predicted performance, if the target performance is not sufficient, the process is stopped at this point, and the design proposal is changed (corrected), and then the tire design is started again. (Redo) or the result of the evaluation of the prediction performance may be stored and referred to as appropriate.

In step 114, the value OBJ _i of the objective function and the value G _{i of the} constraint after the design variables are changed by Δr _{i with} respect to the tire correction model obtained in step 110 are calculated. The sensitivity dOBJ / dr _i of the objective function, which is the ratio of the change amount of the objective function to the unit change amount of the design variable, and the sensitivity dG / dr _i of the constraint condition, which is the ratio of the change amount of the constraint condition to the unit change amount of the design variable. Is calculated for each design variable.

[0103]

(Equation 2)

With this sensitivity, it is possible to predict how much the value of the objective function and the value of the constraint condition change when the design variable is changed by Δr _i . This sensitivity is
Depending on the method used for modeling the tire and the nature of the design variables, the value may be obtained analytically. In this case, the calculation in step 114 becomes unnecessary.

In the next step 118, using the initial value OBJo of the objective function, the initial value Go of the constraint, the initial value ro of the design variable, and the sensitivity, the objective function is maximized while satisfying the constraint by mathematical programming. Predict changes in design variables. Step 12 is performed using the predicted values of the design variables.
At 0, a tire correction model is determined by the same method as in step 110, and an objective function value is calculated.

When the tire correction model is determined in step 120, in the next step 122, the tire performance prediction processing of the corrected tire model is executed in the same manner as in steps 104 and 112. When the tire performance prediction processing is executed in step 122, it is possible to determine whether or not the prediction performance is good from the evaluation of the prediction result in the same manner as described above. This determination may be made by input from a keyboard.Also, an allowable range is predetermined for the evaluation value, and when the evaluation value of the prediction result is within the allowable range, it is determined that the prediction performance is good. You may make it. Also, as a result of the evaluation of the predicted performance, if the target performance is not sufficient, the process is stopped at this point, and the design proposal is changed (corrected), and then the tire design is started again. (Redo) or the result of the evaluation of the prediction performance may be stored and referred to as appropriate.

In the next step 124, step 120
It is determined whether the value of the objective function has converged by comparing the difference between the objective function value OBJ calculated in step (1) and the initial value OBJo of the objective function calculated in step 104 with a threshold value input in advance. If the value of the objective function has not converged, steps 106 to 124 are repeatedly executed with the design variable value obtained in step 118 as an initial value. When it is determined that the value of the objective function has converged, the value of the design variable at this time is used as the design variable value that maximizes the objective function while satisfying the constraints while considering the tire performance. Is used to determine the shape of the tire.

As described above, in the present embodiment, while simulating a tire model and a road surface state, design variables that give the optimum value of the objective function satisfying the constraint conditions are obtained, and the tire is designed from the design variables. Therefore, when designing and developing, unlike conventional design and development based on trial and error, it is possible to perform from the best mode design to the performance evaluation of the designed tire mainly by computer calculation, and achieve significant efficiency improvement. Thus, development costs can be reduced.

In particular, in the case of a two-wheeled vehicle tire, the linearity of the camber thrust with respect to the camber angle is important for improving the steering stability from straight running to turning. However, when the shape of the outer surface of the tread is approximated by one or several curvatures, when the small camber angle changes at any camber angle, the change rate of the camber angle becomes discontinuous. This phenomenon affected the camber thrust at the discontinuous camber angle, causing the linearity of the camber thrust to be impaired. Therefore, when the camber angle is slightly changed, the linearity of the camber thrust can be improved by approximating the outer surface shape of the tread so that the change rate becomes smoothly continuous.

The production of a tire having the above performance is as follows.
By relying on the experience of the designer or by using simple mathematical formulas, the idea was to make a prototype of the tire, test it, and repeat the improvement. In this embodiment, the FEM
By using the optimization method, it is possible to reduce the number of prototypes and improve the development efficiency.

Further, in these optimization methods used in tire development, the calculation time of the FEM portion was enormous, but by using the above method, optimization in a shorter time became possible. That is, when calculating while shifting the tire model from the stationary state to the rolling state, an enormous calculation load is required for the FEM calculation.
For this reason, in the present embodiment, the initial state can be started as a rolling state of an arbitrary angular velocity by giving an angular velocity to the tire model as an initial condition, and the load on the FEM portion where the calculation time becomes enormous in the optimization method is increased. Reduction can be performed and optimization can be performed in a shorter time.

FIG. 18 shows the tire outer shape at the initial stage of optimization and the tire outer shape after optimization according to the present embodiment. It can be seen that the tread has a slightly bulged shape in the middle abdomen. Also, FIG.
The evaluation results of the actual evaluation of the tire No. were shown. FIG. 19 shows the camber thrust when the camber angle is changed. As understood from FIG.
In the initial stage tire, the characteristic that the camber angle is largely bent at around 20 degrees, after the optimization of the present embodiment,
The bending is alleviated, and the linearity is improved. In addition,
The following table shows the results of measurement of various characteristics of each of the tires at the initial stage of optimization and the tires after optimization according to the present embodiment, which were mounted on an actual vehicle. It is understood from the following table that the characteristics of the tire after optimization are improved.

[0113]

[Table 1]

In the above embodiment, curve approximation, Lagrange interpolation, and the like are used as a method of expressing the shape of a line. In addition, circular interpolation, MATHEMATICAL ELEMEME, etc.
NTSFOR COMPUTER GRAPHICS (David F. Rogers and J. Ala
n Adams), a spline curve, a B-spline curve, a Bezier curve, or a NURBS (weighted B-spline) may be used for interpolation.

In the above-described embodiment, the case where the tire is designed by using the shape of the tread line as a design variable has been described. However, the present invention is not limited to this, and the carcass is formed by using the nodes on the carcass line. The shape of the line may be changed. In addition to the shape of the tread line, the shape of the carcass line, the ply line, the line representing the outer shape of the tire, the line of the reinforcing material, and the like may be used as the design variables.

The present embodiment can be applied to the determination of the shape of a plurality of lines. That is, it is possible to determine the shape of the tread tine (outer surface shape), the shape of the carcass line, and the shape of the ply line that improve the steering stability without impairing the riding comfort of the vehicle. For example, in this case, the lateral spring constant can be used as a physical quantity for improving the steering stability as the objective function, and the condition that the upper and lower first-order eigenvalues, which are the physical quantities governing the riding comfort, are constant can be adopted as the constraint condition. it can. Then, the shape of the carcass line, the shape of the ply line, and the shape of the tire outer surface that maximize the lateral spring constant under the condition that the upper and lower primary eigenvalues are constant may be determined. The design variables are
A ply line and a line representing the outer shape of the tire may be employed. The number of lines used as design variables is
A plurality, ie, two or more lines, may be adopted as design variables.

The determination of the shape of the plurality of lines can be applied to the determination of the shape of the bead filler of the tire and members around the bead filler. That is, it is possible to determine a bead filler shape and a rubber chafer shape that reduce rolling resistance without impairing the durability of the bead portion. In this case, it is preferable that the objective function is set to a rolling resistance value and the constraint condition is set such that the principal strain under load at the end of the folded ply is within + 3% of the initial structure. As the design variables, it is preferable to use a line that defines the outer shape (gauge distribution) of the bead filler or a boundary line between the rubber chafer and the side rubber.

The determination of the shape of the plurality of lines can be applied to the determination of the thickness between the respective belt layers in the belt portion. Can be determined. In this case, the objective function is defined as the rolling resistance value, and the constraint condition is that the principal strain under load at the end of the belt end ply is + 3% of the initial structure.
It is preferable to determine within. In addition, a line representing each belt layer is used as a design variable, and one or a plurality of lines change, so that a gauge distribution between each belt layer that optimizes the objective function can be determined.

Furthermore, the present invention can be applied to a belt structure. In this case, the durability can be improved by minimizing the strain concentration occurring in the belt portion without increasing the weight of the belt portion. In this case, it is preferable that the objective function is set to the maximum value of the principal strain under load generated between the belt layers and the constraint is set so that the total belt weight is within + 1% of the total weight of the initial structure. Further, the design variables can determine the angle, the driving (for example, the number of drivings and the driving strength) and the width of each belt layer.

The present invention can be applied to determining the shape of a tire crown. In this case, the pressure distribution in the contact area can be made uniform without changing the shape of the contact area between the tire and the ground, and the wear performance can be improved.
In this case, the objective function is the standard deviation of the pressure distribution in the contact area, and the constraint conditions are the contact length in the tire circumferential direction at the center of the crown and at the end of the belt at the contact length in the initial shape ±
It is preferable to set it within 5%. Further, it is preferable that the design variable is determined as a crown shape by a Lagrange interpolation routine or the like.

In another application of determining the shape of the tire crown, the objective function is the standard deviation of the pressure distribution in the contact area, and the constraint is the contact length in the tire circumferential direction at the center of the crown and at the belt end. ± of length
With respect to the crown shape, which is determined within 5% and is a design variable, it is preferable that the range of the crown portion specified in advance is approximated by a plurality of arcs.

The present invention can be applied to the determination of the shape of the tire pattern surface. In this case, the pressure distribution when the pattern comes into contact with the ground can be made uniform, and the wear performance can be improved. It is preferable that the objective function is set to the standard deviation of the pressure distribution in the contact area, and the constraint condition is set so that the total volume of the pattern is within ± 5% of the initial volume. The shape of the pattern surface, which is a design variable, may be obtained by dividing the pattern surface into a lattice shape in accordance with a previously input Lagrange interpolation order, and using the obtained coordinate values in the pattern thickness direction of each point as design variables. preferable.

It is not necessary to carry out each of the above individually. For example, each of ply line determination and belt structure determination is performed simultaneously, or crown shape is determined using ply lines. They may be combined.

[Second Embodiment] Next, a second embodiment will be described. In the above-described embodiment, a case has been described in which the tire performance prediction and evaluation are repeated for one design proposal while modifying the design proposal, and a design proposal to be adopted is obtained. However, this embodiment is based on a plurality of design proposals. This is a request for a design plan to be adopted. Specifically, in order to improve steering stability, the shape of the tread line that maximizes the camber thrust, which is the optimum value, is genetically designed by an algorithm. In this embodiment, since the configuration is the same as that of the above-described embodiment, the same portions are denoted by the same reference numerals and detailed description is omitted.

FIG. 20 shows a processing routine of a program according to the second embodiment. In step 300,
The N tire cross-sectional shapes are modeled by a method capable of numerically and analytically obtaining the tire circumferential belt tension at the time of air filling, such as the finite element method, and a tire basic model including the internal structure is obtained. The user inputs N in advance. The basic tire model used in the present embodiment is the same as that shown in FIG. 9, and the division of the basic tire model is the same as in the embodiment.

In the next step 302, an objective function representing a physical quantity for evaluating tire performance, constraints for restricting the tire cross-sectional shape, and design variables for determining the tire cross-sectional shapes of the N tire models are determined. Note that the objective function OBJ and the constraint condition G are the same as those in the above-described embodiment, and a description thereof will be omitted. The design variables are determined for each of the N tire models by the curve approximation described above.

As described above, the objective function OBJ, the constraint condition G, and the design variables r _iJ (J
= 1, 2,..., N).
3, for each of the N tire models of the initial generation, in the same manner as in step 104 of FIG. 2 described above,
A tire performance prediction process is executed. Step 30
In No. 3, when the tire performance prediction process is executed, it is possible to determine whether or not the prediction performance is good from the evaluation of the prediction result. Can be stored and can be referred to as appropriate.

FIG. 2 after the tire performance prediction processing is completed.
In step 304 of 0, each objective function OBJ _J and constraint condition G _J of each design variable r _iJ of each of the N tire models
Is calculated. In the next step 306, step 304
Objective function OBJ _J for each of the N tire models determined by
Then, the adaptive function F _J of each of the N tire models is calculated according to the following equation (4) using the constraint conditions G _J. In the present embodiment, for example, in order to maximize the camber thrust, the value (fitness) based on the adaptive function increases as the camber thrust increases.

Φ _J = −OBJ _J + γ · max (G _J , O) F _J = −Φ _J (4) or F _J = 1 / Φ _J or F _J = −a · Φ _J + b

[0130]

(Equation 3) c: constant γ: penalty coefficient Φ _min = min (Φ ₁ , Φ ₂ ,... Φ _N ) Φ _J : penalty function of the J-th tire model of N tire models (J = 1, 2, 3,. .. N) Note that the user inputs c and γ in advance.

In the next step 308, two models to be crossed are selected from the N models. As a selection method, a generally known fitness proportional strategy is used, and the probability P _l that an individual l of N tire models is selected by selection is represented by the following equation.

[0132]

(Equation 4) Here, F _l : an individual l among N tire models
F _J : J-th adaptive function of N tire models J = 1, 2, 3,... N In the above embodiment, the fitness proportional strategy was used as the selection method. Genetic algorithm (Hiroaki Kitano
Edition), expectation strategy, rank strategy, elite preservation strategy, tournament selection strategy, or GE
A NITOR algorithm or the like may be used.

In the next step 310, it is determined whether or not the two selected tire models are crossed according to the probability T input by the user in advance. As used herein, the term “crossover” refers to exchanging some of the elements of the two tire models. If no crossover is determined in the negative determination, the process proceeds to step 316 in step 312 with the current two tire models as they are. On the other hand, when the vehicle is crossed in the affirmative determination, the two tire models are crossed in step 314 as described later.

The crossing of the two tire models is performed by a crossing routine shown in FIG. First, the two tire models selected in step 308 of FIG. 20 are referred to as a tire model a and a tire model b, and the design variables of each of the tire models a and b are represented by a design variable vector including a sequence. Vr the design variable vector of _{^{a a = (r 1 a,}} r 2 a, ···, r i a, ··
^{_{·, R n-1 a)}} , Vr a design variable vector of the tire model _{^{b b = (r 1 b,}} r 2 b, ··· r i b, ··· r
_n-1 ^b ).

In step 350 of FIG. 21, a predetermined random number is generated, and an intersection point i of the tire model a, b with respect to the design variable vector is determined according to the random number. In the next step 352, cross the determined tire model a, b design variables r _i ^a, the relative r _i ^b, determine the distance d according to the following equation. ^{_{d = | r i a -r i}} b |

In the next step 354, r _i ^a , r _i ^b
With the possible range of the minimum value B _L and the maximum value B _u in,
The normalized distance d 'is obtained according to the following equation.

[0137]

(Equation 5)

In step 356, in order to appropriately disperse the value of the normalized distance d ', a mountain-shaped mapping function Z (x) (0≤x≤1) as shown in FIGS. , 0 ≦ Z
(X) ≦ 0.5) and the function value Z according to the following equation:
_Ask for _ab . Zab = Z (d ') in this way, after obtaining the function values Z _ab, Step 35
The new design variables r _i ^'a, r _i' in 8 Request ^b according to the following equation.

[0139]

(Equation 6)

After obtaining r _i ′ ^a and r _i ′ ^b in this manner, design variable vectors Vr ′ ^a , Vr ′ ^b , which are new arrangements of design variables, are obtained in step 360 as follows. ^{_{Vr 'a = (r 1 a}} , r 2 a, ··· r i' a, r i + 1
_{^{b, ···, r n-1}} b) Vr 'b = (r 1 b, r 2 b, ··· r i' b, r i + 1
^a , ..., r _n-1 ^a )

The minimum value B _L and maximum value Bu of the possible range of r _i are input by the user in advance. Further, the mapping function Z (x) is as shown in FIGS. 23 (a) and 23 (b).
It may be a valley-shaped function. Also, in the above example, the crossover location i is 1
In addition to the above, a multi-point crossover or a uniform crossover as shown in a genetic algorithm (edited by Hiroaki Kitano) may be used.

After generating two new tire models by such crossover, in step 316 of FIG.
At the probability S input in advance by the user, it is determined whether or not to mutate. As will be described later, this mutation means to change a part of the design variables minutely, and to increase the accuracy including a population that can be an optimal design variable. If it is determined in step 316 that the mutation is not to be performed, the process proceeds to step 321 while maintaining the current two tire models in step 318. When the mutation is performed in the affirmative determination, the mutation is performed in the following step 320 as follows.

This mutation is performed by the mutation routine shown in FIG. First, in step 362, a random number is generated, and the location i of the mutation is determined by the random number. In the next step 364, the distance d 'is determined by a random number in the range of 0≤d'≤1.

In the next step 366, FIG.
A mountain-shaped mapping function Z (x) (0 ≦ x ≦
1, 0 ≦ Z (x) ≦ 0.5) or FIG.
Using a valley-shaped mapping function Z (x) as shown in FIG.
The function value Zd is obtained according to the following equation. Zd = Z (d ')

After obtaining the function value Zd in this way,
In step 368, a new design variable r _i ′ is obtained according to the following equation.

[0146]

(Equation 7)

After the design variables r _i ′ have been obtained in this manner, the design variable vector Vr ′, which is a sequence of new design variables, obtained in step 370 is as follows. _{Vr '= (r 1, r} 2, ··· r i', r i + 1, ··
·, R _n-1 )

In the next step 321 for the two newly generated diamond models in the same manner as in step 303 (step 10 in FIG. 2).
4) A tire performance prediction process is executed for each tire model. In step 321, when the tire performance prediction process is executed, it is possible to determine whether or not the prediction performance is good from the evaluation of the prediction result. Here, the output of the prediction result and the prediction performance The configuration is such that the result of the evaluation can be stored and can be referred to as appropriate.

In the next step 322, the value of the objective function and the value of the constraint condition are calculated for the two newly generated diamond models. In the next step 324, an adaptive function is calculated from the obtained value of the objective function and the value of the constraint condition by using the equation (4) in the same manner as in the embodiment.

In the next step 326, the two tire models are stored. In the next step 328, it is determined whether or not the number of tire models stored in step 326 has reached N. If the number has not reached N, steps 308 to 328 are repeated until the number becomes N. Execute. On the other hand, if the number of tire models reaches N, convergence determination is made in step 330, and if not, the N tire models are updated to the tire models stored in step 326, Steps 308 to 330 are repeatedly executed. On the other hand, if it is determined in step 330 that the convergence has been achieved, the constraint condition is substantially satisfied with the design variable value of the tire model that maximizes the objective function value while substantially satisfying the constraint condition among the N tire models. In step 332, the shape of the tire is determined using the value of the design variable while maximizing the objective function.

The convergence determination in step 330 is regarded as convergence if any of the following conditions is satisfied. 1) The number of generations has reached M 2) The number of line strings with the largest value of the objective function has become q% or more of the whole 3) The value of the largest objective function is updated in the next p generations Not done.

The user inputs M, q, and p in advance. The output of the tire performance can be used for the convergence determination.

As described above, in this embodiment, since the amount of calculation is increased as compared with the first embodiment, the time required for the design and development slightly increases, but it is necessary to design a tire having better performance. There is an effect that can be.

By applying the present embodiment to the belt structure, it is possible to determine a belt structure that is optimal in terms of opposing performances for improving the steering stability without impairing the riding comfort of the vehicle. In this case, it can be realized by making the selection of the objective function, the constraint condition and the design variable, the crossover method, and the mutation method different. For example, a lateral spring constant, which is a physical quantity for improving steering stability, is used as an objective function, and a condition in which a vertical spring constant, which is a physical quantity that governs ride comfort, is used as a constraint, and a constant vertical spring constant is used. Under the conditions, the belt structure that maximizes the lateral spring constant can be determined.

Although the belt structure is used as a design variable, other reinforcing materials and the like may be determined as design variables. Further, the present invention can be applied to a case where a reinforcing material or the like is inserted into the bead portion.

Further, the present invention can be applied to determination of a block shape of a block-formed tire. In this case, the uneven wear performance can be improved by reducing the block rigidity difference between blocks of different sizes arranged in the tire circumferential direction and making the block rigidity in each direction uniform in one block.

In this case, it is preferable to use uniform block stiffness in each direction as the objective function representing the physical quantity for evaluating the performance of the block. The constraints include the area of the block, the sipe length, and the sipe length × sipe length. There are restrictions on the depth, number of sipes, etc., the amount of change in the sipe length is within ± 5% of the sipe length in the initial model, and the node coordinates of the sipe are in the figure surrounded by the node coordinates of the block The distance between the sipe node coordinates and the line representing the outside of the block is 2mm
It is preferable to determine the above. The design variables are related to the node coordinates of the block, the node coordinates of the sipe, the groove angle of each side of the block, the groove depth of each side of the block, the sipe width, the sipe embedding angle, and the sipe depth. It is preferable to use sipe node coordinates (block node coordinates are fixed).

Note that the first embodiment and the second embodiment may be combined. In other words, when calculating an objective function and constraints based on a design plan obtained by crossover and mutation, Goldberg, DE, "Genetic Algorithms
in Search, Optimization and Machine Learning i ", Addis
As described in on-Wesley (1989), there is a problem that it is difficult to find a true optimal solution although it does not fall into a local optimal solution. Therefore, the above problem can be solved by combining the methods using the processing of steps 104 to 116 of the first embodiment as the arithmetic processing of step 222 of the second embodiment. Such a method of obtaining a true optimal solution without falling into a local optimal solution is to combine a method called Simulated Annealing described in the above-mentioned reference in addition to the method described here. Can also.

[Third Embodiment] In the present embodiment, the present invention is applied to an optimizing device for obtaining optimum tire design parameters. In the optimizing device of the present embodiment, a design parameter is obtained by optimization calculation using a neural network after learning, which is a nonlinear prediction technique in which a neural network of a higher animal is modeled by engineering, as a conversion system. In this embodiment, since the configuration is substantially the same as that of the above-described embodiment, the same portions are denoted by the same reference numerals and detailed description thereof will be omitted.

The optimizing device 60 for performing the optimization according to the present embodiment can be embodied in the same configuration as the personal computer for practicing the pneumatic tire designing method shown in FIG. Predicts the performance of tires from the shape, structure, pattern design parameters, etc. using a neural network based on a nonlinear prediction method according to a program stored in advance and satisfies constraints and optimizes an objective function (for example, maximum or minimum). ) Is calculated. CRT14. The calculation result of the computer main body 12 is displayed.

More specifically, as shown in FIG. 25, the optimizing device 60 includes a computer main body 12 including a microcomputer, a data input / output device 28, a keyboard 10 for inputting data and commands, and a monitor. 14. The computer body 12 is C
PU 16, ROM 18, RAM 50, memory 52 for storing a conversion system and the like (details will be described later), and input / output device (hereinafter referred to as I / O) 56 for exchanging data and the like between the main body and another device. And a bus 54 connected to these so that data and commands can be input and output. The ROM 18 stores a processing program described later. The data input / output device 28
When the numerically expressed tire shape, structure, pattern design parameters, manufacturing conditions, and tire performance (in this embodiment, tire shape, structure, pattern, etc.) are stored in the external storage means, the external storage means And is unnecessary when the keyboard 10 is used as an input device.

FIG. 26 is a block diagram showing a schematic configuration for each function of the optimizing device 60 according to the present embodiment. The optimization device 60 of the present embodiment optimizes the tire performance to be maximized or minimized (this is called an objective function) and outputs design parameters for the optimized tire performance.

The optimizing device 60 includes, for each function, a nonlinear operation unit 62, an optimization operation unit 64, a performance prediction data input unit 70, an optimization item input unit 72, and an optimization result output unit 74.
are categorized. The non-linear operation unit 62 functions as a calculation unit of a conversion system formed by a neural network (details will be described later), and based on the data input from the performance prediction data input unit 70, the shape, structure, pattern, and manufacturing conditions of the tire. And a conversion system in which the performance of the conversion system is associated. Here, the conversion system means a conversion system itself that can perform conversion and inverse conversion such that the shape, structure, pattern design parameters, manufacturing conditions, and the like of the tire correspond to the performance one-to-one. When the neural network after learning is expressed by a mathematical expression, it refers to the expression including the mathematical expression and its coefficient. The performance prediction data input unit 70 is for inputting data such as tire shape, structure, pattern design parameters, manufacturing conditions, and the like, and performance corresponding thereto. The performance in this case is obtained by the tire performance prediction processing.

The optimization item input section 72 predicts a tire to be maximized or minimized or measures tire performance such as measured physical quantity (objective function described later), or predicts or estimates a tire to be restricted when maximizing or minimizing. Physical parameters to be measured, design parameters for tire shape, structure, and pattern, as well as manufacturing conditions such as vulcanization temperature, possible ranges of design parameters and manufacturing conditions for tire shape, structure, and pattern, and methods for optimization. This is for inputting selection and parameters at that time.

The above-mentioned optimization methods include optimization methods such as mathematical programming and genetic algorithms. In the present embodiment, it is assumed that an optimization method based on mathematical programming is selected.

The optimization operation section 64 is for optimizing the objective function until it converges, and is composed of an objective function / constraint condition operation section 66 and an objective function optimization operation section 68. The objective function / constraint condition calculation unit 66 is a non-linear calculation unit 62
Is used to predict the tire performance from the design parameters and manufacturing conditions of the shape, structure, and pattern of the tire by using the conversion system according to the above. This is to optimize the function until it converges while satisfying the constraints.

The optimization result output unit 74 includes the optimization operation unit 6
As a result of the optimization according to No. 4, the tire shape, structure, pattern design parameters and manufacturing conditions optimized to satisfy the input optimization item are output.

In the present embodiment, the nonlinear operation unit 6
2 is configured using the hardware resources shown in FIG. 25 and software resources to be described later, has a conversion function formed by a conceptual neural network as described later, and has a learning function to learn the same. I have. Further, the non-linear operation unit 62 may be configured to have only a conversion function without a learning function. That is,
As will be described later, the non-linear operation unit 62 determines the shape of the tire,
A conversion system in which the design parameters and manufacturing conditions of the structure and the pattern and the tire performance are associated with each other is obtained. However, it is sufficient that the design parameters and the manufacturing conditions of the tire shape, structure and pattern and the manufacturing condition and the performance can be converted. . Therefore, the correspondence between the design parameters and the manufacturing conditions of the tire shape, structure, and pattern and the performance thereof is learned in advance by another neural network, and the learned conversion coefficient of the other neural network is input. The conversion system in which the design parameters of the tire, the design parameters of the pattern, and the manufacturing conditions are associated with the performance using the coefficients may be obtained. That is, as long as the conversion coefficient is input, a function of only conversion for converting between the tire shape, structure, pattern design parameters and manufacturing conditions and the tire performance using the conversion coefficient may be used. Alternatively, these correspondences may be stored as a look-up table and converted by referring to the stored look-up table.

The above-mentioned non-linear operation section 62 is used as an input layer to input design parameters of tire shape, structure and pattern and design parameters of tire shape, structure and pattern. It has neurons according to the number of manufacturing conditions, and as output layers via the intermediate layer, neurons according to the number of tire performance items to be predicted related to objective functions and constraints, and each neuron is connected by synapses. Of the neural network. When the values of the design parameters and the manufacturing conditions of the shape, structure, and pattern of the tire are input after learning, which will be described later, the non-linear operation unit 62 outputs the performance corresponding thereto. At the time of learning, the known performance corresponding to the design parameters of the tire shape / structure / pattern and the manufacturing conditions are input as a teacher, and the shape / structure / tire / size of the tire is determined by the magnitude of the error difference between the output performance and the known performance. Each value of the pattern design parameter and the manufacturing condition is set so as to correspond to the performance.

As an example of the neural network used in the non-linear operation unit 62, as shown in FIG. 27, a predetermined number of units I1, I
, Ip (p> 1), an intermediate layer composed of a number of units M1, M2,..., Mq (q> 1), and a predetermined number of output units U1, U2,.・ ・ 、
It is composed of an output layer made of Ur (r> 1). The number of units in the input layer and the number of units in the output layer may be set in accordance with the design parameters of tire shape, structure, and pattern, the number of manufacturing conditions, and the number of performances. Also,
Each unit of the intermediate layer and each unit of the output layer are connected to offset units 46 and 48 for offsetting the output value by a predetermined value. For the unit of the input layer, for example, a parameter representing the width of the tire belt, the angle of the belt, the material of the belt, the shape of the tire, and the cost can be used as input values. For the output layer unit, for example, rolling resistance, stress / strain, tire spring characteristics, tire grounding characteristics, and the like can be used as output values.

In this embodiment, the unit of the intermediate layer and the unit of the output layer are constituted by neural circuit elements having sigmoid characteristics whose input / output relation is represented by a sigmoid function, and the unit of the input layer is connected to the input / output relation. Are composed of linear neural circuit elements. By configuring so as to have this sigmoid characteristic, the output value becomes a real value (positive number).

The outputs of the units of the intermediate layer and the units of the output layer in the nonlinear operation unit 62 can be expressed by the following equations (5) and (6). That is, for a unit, the number of synapses on the input side is p,
Assuming that a weight (coupling coefficient of a unit) corresponding to the strength of each synaptic connection is w _ji (1 ≦ j ≦ N, 1 ≦ i ≦ p) and each input signal is x _j , an average of the membrane potential of the neuron The virtual internal state variable u corresponding to the value can be expressed by the following (5), and the output y can be expressed by the following equation (6) using a nonlinear function f indicating the characteristics of the neuron.

[0173]

(Equation 8)

Here, b _j represents the offset value supplied from the offset unit, and W _ji represents the weight between the i-th and j-th units in different layers.

Therefore, by inputting each value of the design parameter and the manufacturing condition of the tire shape, structure, and pattern to the unit of the input layer, each value corresponding to the number of tire performances is output from the unit of the output layer. You.

Note that the characteristics of each unit in the input layer described above may be characteristics that output the input as it is. Further, the weight (coupling coefficient) of each unit of the non-linear operation unit 62 (neural network) is learned / corrected by a learning process described later so that the error of the known prediction data is minimized.

Next, the details of the neural network learning process in the nonlinear operation section 62 will be described with reference to FIG. In the present embodiment, data relating to the performance of the tire is obtained by trial production / evaluation of the tire based on the values of the design parameters and manufacturing conditions of the shape, structure, and pattern of the tire, or by modeling and predicting the tire by a computer. In addition, the correspondence between each value of the design parameters and manufacturing conditions of the shape, structure, and pattern of the tire and each value representing its performance is used as learning data.

A predetermined number (for example, 90% of the entire data) of a plurality of data is used as learning data, and the other data (for example, the remaining 10%) is used as test data. This is because the prediction data is used for data to be used at the time of learning of the neural network and data for confirming whether or not the neural network after the learning has been optimally learned. Each value of the design parameters and manufacturing conditions of the shape, structure, and pattern of the tire is used as input data, and each value representing the performance of the tire is used as output teacher data.

First, in step 460 of FIG. 28, the learning data and the test data obtained in advance are read. In the next step 462, initialization is performed by setting the coupling coefficient (weight) and offset value of each unit in the neural network to predetermined values. In the next step 464, an error of each unit of the intermediate layer and the output layer is determined in order to train the neural network using a plurality of learning data whose design parameters and manufacturing conditions of the shape, structure, and pattern of the tire are known.

The error of the output layer can be the difference between the learning data and the tire performance. By slightly changing at least one of each coupling coefficient and offset value, the error of the output layer, that is, the error of the unit can be minimized. The error in the intermediate layer can be obtained by an inverse calculation such as an error back propagation method using the error in the output layer.

In the next step 466, each of the obtained coupling coefficients and offset values is updated (rewritten), and in the next step 468, each of the test data is converted by the neural network based on the updated coupling coefficients and offset values. The test is performed to obtain data representing the performance of the tire as a value of the test result. In the next step 470, it is determined whether or not the value of the test result obtained in the above step 468 is within a predetermined range which is a criterion of convergence determination, thereby determining whether or not the convergence is achieved. Is repeated a predetermined number of times, and in the case of an affirmative determination, this routine ends. On the other hand, if a negative determination is made, the process returns to step 464 and the above processing is repeated. Thus, when learning data is input, each coupling coefficient and offset value are determined so that the error of each unit of the intermediate layer and the output layer is minimized.

As described above, the neural network is trained by using a plurality of prediction data whose tire shape, structure, pattern design parameters and manufacturing conditions are known. That is, learning is performed so that the error of the output value from the output layer of the neural network with respect to the teacher signal is minimized. As described above, by learning, when the values of the design parameters of the shape, structure, and pattern of the tire and the values of the manufacturing conditions are input, the nonlinear operation unit 32 outputs a value representing the performance of the tire.

After the above processing is completed and the learning of the neural network is sufficiently performed, the network structure, that is, the coupling coefficient and the offset value are stored in the memory 52.
And a conversion system may be constructed.

In the above, the case where a neural network is used as the non-linear operation section 62 has been described.
As shown in the equation, a conversion system using a response surface method based on a polynomial can also be used.

[0185]

(Equation 9)

Next, the operation of the optimizing device 60 according to the present embodiment will be further described with reference to the flowchart in FIG. When the power of the optimizing device 60 is turned on or an instruction to start execution is given from the keyboard, step 400 in FIG.
Then, the design parameters x _i (i = 1 to p) of the tire shape, structure, and pattern, the objective function, and the maximum number of experiments are set. That is, which performance is to be improved, and in that case, how many times the number of experiments is desired to determine the optimal tire shape, structure, and pattern design parameters are determined.

In the next step 402, the step 400 is executed.
Design parameters for tire shape, structure and pattern set in
Data x_iSet the tolerance of (x_i ^L≤x_i≤x_i ^U:
x_i ^LIs the lower limit, x_i ^UIs the upper limit), the next step 40
In the fourth section, the number of analysis M and the tires by experiment or numerical calculation
Represents the position of the design parameters of the shape, structure, and pattern
The variable e is initialized (M = 0, i = 1).

In the next step 406, step 400
It is determined whether or not past experimental data can be used with respect to the tire shape, structure, pattern design parameters x _i , and tire performance set in the step. If it is necessary to obtain the value, go to step 420.

[0189] At step 420, using the orthogonal array or optimal experimental design, etc., the shape of any of the tire structure, shape of the tire by determining whether to perform experiments by changing the design parameters x _i of pattern, structure And the design parameters of the pattern are determined. The determination of the design parameters for the shape, structure and pattern of this tire is described in Boxa
nd Draper; "Empirical Model Builing and Response Su
rfaces ", John Wiley & Sons, New York".

In the next step 422, step 420
Tire shape, structure, and pattern according to the experimental plan determined in
The design parameters for each parameter are changed according to the
Create an ear model. That is, the total number of experiments or numerical values
Number of analysis is n_iAs n _iTire cross section
Tension in the tire circumferential direction at the time of air filling as in the limiting element method
The method is capable of calculating the force numerically and analytically.
Delling and obtaining a tire basic model including the internal structure.

In step 422, an objective function representing a physical quantity for tire performance evaluation, constraints for restricting the tire cross-sectional shape, and design variables for determining the tire cross-sectional shapes of the n _i tire models are determined. Thereafter, in step 424, for each of the n _i tire models, in the same manner as in step 104 of FIG. 2 described above,
Execute tire performance prediction processing. As a result, a tire performance prediction result is obtained.

FIG. 2 after the tire performance prediction processing is completed.
In step 426 of step 9, the objective function OBJ _J and the constraint condition G _J of each of the design variables of the n _i tire models are calculated and stored.

At the next step 428, the neural network is learned as described above. That is, the neural network learns the values input to the input layer as the values of the design parameters of the shape, structure, and pattern of the tire, and the values output from the output layer as the values of the performance of the tire.

In the next step 430, it is determined whether or not there are tire shape, structure, and pattern design parameters that have little contribution to the target physical properties and characteristics. For example, the sensitivity showing a changing trend in the tire performance of the output layer for the time obtained by slightly changing the shape of at least one of the tire enter into the unit, structure, pattern design parameter x _i of the input layer,
And calculating the degree of decrease in the prediction accuracy of the tire performance of the output layer with respect to the case where the output from at least one unit of the input layer is made zero, and the shape, structure,
Determine the design parameters for the pattern. This is because the design parameters of the tire shape, structure, and pattern, which have low sensitivity and do not reduce the prediction accuracy even if the input is ignored, are considered to have little contribution.

[0195] The shape of less contributing tire, when the structure, design parameters of the pattern is present, an affirmative determination in step 430, deletes the shape of less contributing tire in a next step 432, the structure, the design parameters x _i of the pattern Then, learning is performed again based on the design parameters of the shape, structure, and pattern of the tire after the deletion (step 428). On the other hand, when there is no design parameter of the shape, structure, or pattern of the tire having a small contribution, a negative determination is made in step 430, and in the next step 434, the input layer (the tire shape,
The relationship between the structure and pattern design parameters) and the output layer (tire performance) is stored. That is, each coupling coefficient and offset value are stored.

At the next step 436, the stored input layer (design parameters of the shape, structure and pattern of the tire)
Determining the shape of the best tire, structure, design parameters x _i of the pattern by optimizing the objective function as described later using the relationship between the output layer (tire performances) and (Figure 30).

When the optimization is completed, the next step 438
Increases the number of experiments or analysis M by (M = M +
n _i ), in the next step 440, it is determined whether M <(the set maximum number of experiments or analysis) or not,
If smaller, the process proceeds to step 442.

In step 442, the variable i is incremented. In the next step 444, the following (8) to (10)
As shown in the equation, the allowable range of the design parameters of the shape, structure, and pattern of the tire is reset, and the process returns to step 420. By repeating this process, the optimal tire shape,
The accuracy of the structure and pattern design parameters x _i ^OPT can be improved. The re-setting of the allowable range in step 444 is performed by narrowing the allowable range of the design parameters of the tire shape, structure, and pattern determined in step 402, and in step 420, the re-experiment point is planned for the narrowed region. Do.

[0199]

(Equation 10)

Here, NN is a coefficient for determining the degree of narrowing the allowable range of the design parameters of the shape, structure and pattern of the tire, and it is desirable to set a value of about 1.5 to 5.

On the other hand, if a negative determination is made in step 440, that is, if the number of experiments or numerical analyzes is larger than the predetermined maximum number of experiments or analysis, the design of tire shape, structure, and pattern finally obtained in step 446 is performed. Output parameters as optimal tire design. Next step 4
At 48, it is determined whether or not there is a similar experiment or numerical analysis in the past experimental data. If a negative determination is made, the performance of the optimal tire design is checked in the next step 450 by the memory 22 or the data input / output device 28. Via an external storage device or the like. It should be noted that the performance of the tire may be obtained by an experiment or a numerical analysis again.

Note that the maximum number of experiments or analysis is
It is a constant determined by the cost of an experiment or a numerical analysis and the time used to find the optimal tire design. Next, if the determination in step 406 is affirmative, in step 408 the past tire shape, structure, pattern design parameters related to each item set in step 400 from the prepared database,
The tire performance is read, and in the next step 410,
The read data is converted by using the following equations (11) to (14) so that the kurtosis and the skewness are reduced.

[0203]

[Equation 11]

In the next step 412, the above step 4
The neural network is learned in the same manner as in step 28, and the learning result is stored in the next step 414 as in step 434. In the next step 416, in order to return to the experimental data (predicted data), the inverse conversion of the conversion in step 410 is performed. In the next step 418, the total number of experiments ni is reset (= 0), and the process proceeds to step 436.

Next, details of the optimization processing in step 436 in FIG. 29 will be described. In step 460 of FIG. 30, the objective function representing the tire performance to be improved, the constraints that restrict the tire performance that should not be degraded when a certain tire performance is improved, and the design parameters of the shape, structure, and pattern of the tire are defined. The design variables to be determined are determined, and the next step 4
At 62, a variable j representing the number of design parameters for the tire shape, structure, and pattern is reset (= 0).

At the next step 464, design parameters of the shape, structure and pattern of the tire to be used as initial values when optimizing are set. The problem of optimizing the design of tire shape, structure, and pattern is that the input values (eg, belt width and accuracy) are plotted on a two-dimensional plane, and the value of the objective function is plotted in the height direction. When grasped, since the design space relating to the performance of the tire has multi-modality, it is necessary to know the solution space of the optimal solution by performing optimization from different initial values. As the initial value, for example, the following equation (15) can be used.

[0207]

(Equation 12)

[0208] _{However, x i (i = 1~p)} : shape of the tire, the structure, the design parameters of the pattern ^{_{x i L ≦ x i ≦ x}} i U: shape of the tire, the structure may take the design parameters of the pattern range k = 0 to Munit Munit: the number of divisions of the allowable range of the design parameters of the shape, structure and pattern of the tire

In the next step 466, step 304 is executed.
The output of the neural network is executed by using the initial tire shape, structure, and pattern design parameters set in the above as input, and the performance of the tire corresponding to the input tire shape, structure, and pattern design parameters is predicted.
The initial values of the objective function and the constraint condition are calculated using the result.

In the next step 468, the tire shape,
Structure, the shape of the tire is set in step 464 to change the design parameters of the pattern, structure, and design parameters x _i of pattern each △ x _i increments varied,
In the next step 470 calculates a value G _i of the value OBJ _i and constraints of the objective function after the design variable was [Delta] x _i changes, the following expression in step 472 (16), according to (17), design variables each design sensitivity dG / dx _i ratio is a constraint variation constraints for unit variation in sensitivity Dobj / dx _i and design variables of the objective function is the ratio of the amount of change in the objective function to a unit variation amount of Calculate for each variable.

[0211]

(Equation 13)

[0212] This sensitivity can be predicted whether a value how much change of the objective function when the design variable was [Delta] x _i changes. This prediction, that is, the process of optimization can be compared to mountain climbing, and predicting a change in the value of the objective function is equivalent to instructing the direction of mountain climbing.

In the next step 474, it is determined whether or not the calculation has been completed for all tire shape, structure, and pattern design parameters.
If the calculation has not been completed for the pattern design parameters, steps 468 to 474 are repeatedly executed.

In the next step 476, the change amount of the design variable which minimizes or maximizes the objective function while satisfying the constraint condition is predicted by mathematical programming using the sensitivity of the objective function and the constraint condition with respect to the design variable. Using the predicted values of the design variables, the design parameters of the shape, structure, and pattern of each tire are modified in step 478, and the objective function value based on the modified design parameters of the shape, structure, and pattern of each tire is calculated. . In the next step 480, the objective function value OBJ calculated in step 488 and step 4
It is determined whether the value of the objective function has converged by comparing the difference between the initial value OBJo of the objective function calculated at 66 and a threshold value input in advance, and the value of the objective function has converged. If not, steps 466 to 480 are repeatedly executed using the design variable value obtained in step 476 as an initial value. When it is determined that the value of the objective function has converged, the value of the design variable at this time is used as the design variable value that optimizes the objective function while satisfying the constraints,
In step 482, the design parameters of the tire shape, structure, and pattern are determined using the values of the design variables,
In the next step 484, the variable j is incremented, and the flow advances to step 486.

In step 486, j is the allowable number of design parameters for the initial tire shape, structure, and pattern: (1+
Munit) ^p is determined, and if not, the process returns to step 464 to change the initial design parameters of the tire shape, structure, and pattern, and repeats steps 464 to 486. .

On the other hand, if a positive determination is made in step 486, the optimal tire design is determined in the next step 488, and this routine ends. The determination of the optimal tire design in step 488 of the present embodiment is determined in consideration of the following two conditions, and the one having a high degree of coincidence with the condition is determined as the optimal tire design.

[Conditions] The objective function OBJ has a small value. (Set the smaller the tire performance selected as the objective function, the better. If the larger the better, add a minus sign to deal with it.) The shape, structure, and pattern of the tire around the obtained optimal solution Even if the design parameters are slightly changed, the objective function and constraints do not change much.

As described above, in the present embodiment,
In order to determine the conversion system, in the non-linear operation unit by the neural network, the correspondence between tire shape, structure, pattern design parameters, manufacturing conditions and tire performance is learned by data from experiments or numerical analysis, It is not necessary to assume a functional type as a means for calculating a conversion system, and a conversion that can find a mutual relationship in which correspondence between tire shape, structure, pattern design parameters and manufacturing conditions and tire performance is associated. The system can be created with high accuracy and low arbitrariness. Also,
By combining the conversion system and the optimization calculation unit, it is possible to output an effective design plan of the effective shape, structure, and pattern of the tire.

The optimization may be performed by a genetic algorithm instead of the sensitivity analysis (FIG. 30) in the above embodiment. Note that the method of the genetic algorithm may execute a process excluding the tire performance prediction process of step 303 and step 321 from the process of FIG.

[Fourth Embodiment] Next, a fourth embodiment will be described. In the present embodiment, in order to improve uneven wear suppression performance and steering stability performance, a tire is designed by designing a tread shape of a rectangular parallelepiped block that makes the contact pressure uniform without reducing the contact area.

FIG. 31 shows a processing routine of the program according to the present embodiment. In step 500, similar to step 100 in FIG. 2, a tire model for transforming a tire design plan (change in tire shape, structure, material, pattern, etc.) into a model for numerical analysis, and tire performance evaluation Initial model creation processing such as creation of a road surface model is performed.

In this embodiment, in order to design a tread shape of a rectangular parallelepiped block for making the contact pressure uniform without reducing the contact area, and to design the tire, in step 500, one block constituting the tire is further processed. Modeling is advanced, and the block shape with a flat tread is used as the reference shape, and this reference shape is modeled by a method that can numerically and analytically determine the response when the road surface is input, such as the finite element method, etc. And obtain a block basic model divided into a plurality of elements by mesh division. The reference shape is not limited to a flat tread and may be any shape. Here, modeling refers to digitizing a pattern shape, a block shape, a structure, a material, and the like into an input data format to a computer program created based on a numerical / analytical method.

FIG. 32 shows an example of a block basic model. The block basic model is divided into a plurality of elements by mesh division, that is, divided into a plurality of elements by a plurality of line segments PL in the drawing. In addition,
Although an example in which the block basic model is divided into a plurality of elements as shown in FIG. 32 has been described above, this division method is arbitrary, and the division width can be changed according to the purpose,
Further, it may be divided into an arbitrary shape such as a triangle. Node D _i in the case of this embodiment for determining the block tread surface shape
(I: number of nodes, i ≧ 1) of, and the coordinates of the design variables r _i in the block height direction (arrow UP direction in FIG. 32).

Next, in the same manner as in steps 102 and 104 of FIG. 2, in step 502, an objective function representing a physical quantity for evaluating tire performance, constraints for restricting the tire cross-sectional shape, and design variables for determining the tire cross-sectional shape. Is determined, and in the next step 504, the initial tire performance is predicted.

In the next step 506, the above step 5
At least one input I _j (j: input number, j ≧ 1) is given to the block basic model modeled at 00.
In the present embodiment, a total of nine inputs I _j are given. As shown in FIG. 33A, a flat push load (for example, a surface pressure of 4 kgf / cm ² ) when a load is applied substantially perpendicularly to the block basic model is set as an input I ₁ . Under this flat push load, as shown in FIG. 33 (B), inputs are made in eight directions from the center of the block basic model in the direction intersecting with the block height direction UP (along the tread surface C) and at equal angles. That is, the shear direction inputs of about 45 degrees pitch (eight directions) are given 1 mm each, and these inputs are input I _{2 to} I ₉ . This 1 mm is equivalent to moving the road surface by 1 mm with respect to the reference block bottom surface. These inputs I
_{1 to} I ₉ correspond to input conditions. The input condition is a condition for defining an input to be given, and the input force (load),
Direction and a combination of multiple inputs.

In this embodiment, a plurality of inputs I ₁ to I ₁
Will be described a case of giving the I _9, is not limited to the number of the input, depending on the purpose, may be a load condition on two or more of the plurality. Also, the number of shear direction inputs is not limited to five, and may be one or more, and the load may be set for each of the shear direction inputs.

In the next step 508, the above step 5 is executed.
06, the ground pressure p _{i, j} with respect to the input I _j given is calculated, and an objective function OBJ representing a pattern performance evaluation physical quantity or a block performance evaluation physical quantity (hereinafter, referred to as a pattern / block performance evaluation physical quantity), A constraint condition G for restricting a pattern / block tread shape is determined. In the present embodiment, the objective function (OBJ and constraint conditions (G1, G2)) is determined as follows, with the aim of improving steering stability and simultaneously suppressing uneven wear.

Objective function OBJ: Standard deviation of contact pressure distribution in contact area Constraint condition G1: Equal to or greater than block basic model of input having the same contact area Constraint condition G2: Maximum block height is the same as block basic model

In the present embodiment, the contact area is provided as the constraint condition G from the viewpoint of attaching importance to the contact property, and the maximum block height is not changed for the purpose of keeping the tire radius unchanged. However, other physical quantities can be used for the constraint condition G according to the purpose, and the present design method is satisfied even if the constraint condition G is singular or plural, or does not use the constraint condition G.

In the above description, the contact pressure with respect to the input is obtained in step 508, but the contact pressure distribution may be obtained when the input is given in step 506.

[0231] In the next step 510, using the ground pressure deviation of each node as sensitivity information, (predicts the amount of change in the shape of a block tread) to predict the amount of change in the design variables r _i. That is, if it is higher (lower) than the average from the contact pressure of each node, the amount of change in the improvement direction of the design variable is predicted in the direction of decreasing (increasing) the block height. Step 51 of the present embodiment
In 0 predicts a variation of the design variable r _i in accordance with the following equation (18).

[0232]

[Equation 14]

The above equation (18) indicates that, when considering a plurality of input conditions, a distance is obtained by multiplying the ratio by a proportional constant α according to the input condition having the largest ratio with the average contact pressure. Only changes. " In the equation (18), when assuming a uniform contact pressure, the tread shape (block height) of a portion having a low contact pressure is increased, and the tread shape (block height) of a portion having a high contact pressure is increased.
Is used to make it possible to change the design variable r _i .

[0234] In the next step 512, the system produces a modified model corresponding to the predicted value of the amount of change of the design variable r _i. In this step 512, a tire model can be formed from the modified model or block and a modified tire model can be generated. In the next step 514, a tire performance prediction process of the corrected tire model is executed in the same manner as in step 104 of FIG. Step 51
In 4, when the tire performance prediction processing is executed, it is possible to determine whether the prediction performance is good or not from the evaluation of the prediction result in the same manner as described above. This determination may be made by input from a keyboard.Also, an allowable range is predetermined for the evaluation value, and when the evaluation value of the prediction result is within the allowable range, it is determined that the prediction performance is good. You may make it. Also, as a result of the evaluation of the predicted performance, if the target performance is not sufficient, the process is stopped at this point, and the design proposal is changed (corrected), and then the tire design is started again. (Redo) or the result of the evaluation of the prediction performance may be stored and referred to as appropriate.

After the tire performance prediction processing is completed, in step 516, the contact pressure, the objective function, and the constraint condition are obtained, and in step 518, the above-described step 51 is performed.
By comparing the value of the objective function OBJ obtained in step 6 with the value of the objective function obtained up to the previous iteration (the value of the objective function obtained in step 516 in the previous iteration), the objective function is obtained. It is determined whether or not the value of has converged. When a convergence value of the objective function that satisfies the constraint condition is obtained, the result is affirmative in step 518, and the process proceeds to step 520. On the other hand, if a convergence value of the objective function that satisfies the constraint condition is not obtained, the result in step 518 is negative, and the process returns to step 506 to repeatedly execute the above processing.

In this embodiment, it is expected that the smaller the standard deviation of the contact pressure distribution in the contact area, which is the objective function, the better the uneven wear suppression performance and the steering stability performance. The value is converged in the direction of decreasing value. Therefore, in the present embodiment, the processing from step 506 to step 518 is repeatedly executed until the minimum value of the objective function OBJ is obtained.

In this embodiment, since there is a constraint condition G, when the minimum value of the objective function OBJ is obtained while satisfying the constraint condition G, iterative processing (steps 506 to 506) is performed.
Step 518) ends. The equation defining the change in shape may be other than equation (18), and various methods are possible. For example, a mathematical expression that multiplies a weight w such that a particular input is emphasized among a plurality of inputs and mixes responses from the respective inputs can be used. The following (1
An example is shown in equation 9).

[0238]

(Equation 15) Where wj is the weight for the input Ij

Although the objective function of the present embodiment employs a measure that the smaller the objective function, the better, the minimum value is obtained.
Depending on the purpose, it is possible to select a better objective function with a larger value or an objective function that optimizes a specific value.

At the next step 520, the block tread shape
Curve the shape. This curve fitting process
Is a predetermined curve for the shape near the grounding edge of the block.
This is a process of making the shape of the radius R equal. Specifically, FIG.
4 (A) and FIG. 34 (B),
From the edge of the block to the block
Length (direction along direction UP and direction opposite to direction UP)
H position and the block in the horizontal direction (intersection of direction UP
Direction) and a curve connecting a position of a predetermined length L
The block shape is made uniform so as to have a diameter R. This is the tread
The whole is the design variable r _iAnd the above step 518
Due to processing, the shape of the area near the contact edge or the entire tread is complicated
In consideration of manufacturing labor and cost,
This is to make the shape pure.

FIG. 35 shows an example of the block shape calculated up to step 518. FIG. 36 shows the arrow B in FIG.
2 shows a perspective view from the direction. FIG. FIG. 37 shows the result of performing the curve fitting processing in step 520. Therefore, the shape of the block shown in FIG. 36 is replaced with the shape of the block shown in FIG. The radius of curvature R arranged at each block end is determined by the least squares approximation to be the most optimal design variable r _i.
Can be obtained. This approximation is performed at several points on the block end section, and the other parts can be determined by interpolating the Lagrange polynomial between the points.

In this embodiment, the radius of curvature R is approximately
Is used, the radius of curvature R may be a polynomial, a piecewise polynomial, a spline, a NURBS, a rational function, or the like. Further, although the portion between the designated cross sections is interpolated by the Lagrange polynomial, another polynomial, a piecewise polynomial, a spline function, a NURBS, a rational function, or the like may be used in the same manner as the radius of curvature R. Instead of specifying a cross section, the tread surface shape itself can be represented by a polynomial interpolated surface, a piecewise polynomial surface, a spline surface, a NURBS surface, or the like. In the approximation, an approximation method other than the least squares approximation may be used. As described above, in the present embodiment, the shape is approximated in consideration of the manufacturing after the design variable giving the optimal objective function value is obtained.

When the above processing is completed, the next step 5
At 22, the tread shape is determined. This step 114
Then, the tire is designed based on the design variable that gives the best objective function value obtained from the above calculation, that is, the tire in which the block having the determined tread shape is arranged. In step 522, a tire vulcanizing mold can be designed instead of the tire design.

In the above description, the case where the shape is approximated in consideration of manufacturing after the design variable giving the optimal objective function value has been obtained has been described. In consideration of the capacity of the computer, the calculation may be performed once or every time. FIG. 38 shows an example of a processing flow in the case where the shape is approximated for each operation. FIG.
The process of step 8 is to perform the process of step 520 in FIG. 31, that is, the process of approximating the shape in consideration of manufacturing after the design variable giving the optimal objective function value is obtained, between steps 512 and 514. Things.

[0245] In the process of FIG. 38, in step 512, to create a modified model corresponding to the predicted value of the amount of change of the design variable r _i, processes curve Fuitto block tread surface shape in the next step 520. Then, in step 514, a tire performance prediction is executed.

Therefore, the convergence of the objective function can be determined with the approximate shape by approximating the shape for each operation. In addition, when performing one approximation to the numerical operation,
In FIG. 38, the determination condition may be applied so that the process of step 520 is performed once for each numerical operation.

In the case where the shape is approximated for each calculation, the function as a constraint condition applied to the design variables is achieved.

On the other hand, it is also possible to apply the optimum shape itself to tire design without using shape approximation. FIG. 39 shows a flowchart when the shape approximation is not performed. FIG.
The process of No. 9 is obtained by deleting the process of Step 520 in FIG. 31, that is, the process of approximating the shape in consideration of manufacturing after the design variable giving the optimal objective function value is obtained.

As described above, in the present embodiment,
Since the pattern and the block tread shape are optimized, it is possible to provide a tire with improved steering stability and uneven wear resistance.

FIG. 40 and FIG. 41 are conceptual diagrams showing how the tires are worn. The solid line is a cross-sectional view before the wear and the broken line is a cross-sectional view near the ground end after the wear. As shown in FIG. 40, in the tire having the conventional shape, the contact edge portion is locally severely worn. On the other hand, as shown in FIG. 41, the tire of this example is in a substantially uniform wear state. Accordingly, it is understood that local uneven wear is suppressed by an appropriate tread shape.

As described above, by optimizing the pattern / block tread shape according to the present embodiment, it is possible to provide a tire having improved steering stability and uneven wear resistance.

[Fifth Embodiment] Next, a fifth embodiment will be described. Note that, in this embodiment, since the configuration is the same as that of the above-described embodiment, the same portions are denoted by the same reference numerals and detailed description is omitted.

If the analysis is performed with patterns all around the tire model, the amount of calculation becomes enormous and the results cannot be obtained easily. Therefore, in the present embodiment, in order to facilitate the design of the tire, the tire is designed by giving a pattern only to a part of the tire model.

Therefore, in the present embodiment, in order to facilitate the tire design, the tire model 30 is based on a smooth tire model whose entire circumference is flat, and it is easy to analyze the stepped portion, the ground contact portion, and the kick-out portion. The analysis is performed by including some necessary patterns in the smooth tire model. In the following description, this analysis is referred to as GL (Global-
Local) analysis.

Next, the GL analysis in the present embodiment will be described. The outline of the GL analysis can be implemented by the following procedures 1 to 4.

<Procedure of GL analysis> Step 1: Prepare a smooth tire model, a pattern model (part) and a belt model to be attached to a pattern Step 2: Rolling analysis of a smooth tire model (global ana
lysys: G analysis) Step 3: Pattern part from the result of the smooth tire model
Calculate the rolling locus of the belt model (same as part of the pattern model) to be attached to (part). Specifically, the displacements of all the nodes of the belt model (shell) during rolling are output (this may be converted to speed and output.
If it can be obtained as it is with software constraints and displacements, that is fine), paste a part of the pattern model on the belt model, and force the node at the node of the belt model (displacement is also possible)
Step 4: Since only part of the pattern portion can be rolled up to Step 3, a road surface model corresponding to the pattern portion is prepared, and only the pattern portion is analyzed (local an
alysys: L analysis).

The details are substantially the same as those of the above embodiment. First, a tire model is created from a tire design plan, and a road surface state is input by selecting a friction coefficient μ together with a road surface model, and setting conditions are set. , Design variables and constraints (100 to 104 in FIG. 2). In this case, the tire model is a smooth tire model, and a pattern model (part) and a belt model of a portion to be attached to the pattern are created. Then, setting conditions are set, and deformation calculation of the tire model is performed. This is a rolling analysis (global analysys: G analysis) of a smooth tire model.

[0258] Next, (106-110 in FIG. 2) obtains the model obtained by the unit amount [Delta] r _i changed design variables, tire performance prediction seeking tire model deformation calculation and giving condition (112 in FIG. 2). That is, the rolling locus of the belt model (same as a part of the pattern model) to be attached to the pattern part (part) is calculated from the result of the smooth tire model. As a result, only the pattern part (part) is rolled, so that a road surface model corresponding to the pattern part is prepared, and only the pattern part is analyzed. This is an analysis of only the pattern part which is a part of the pattern model (local analysis: L analysis).

Then, the sensitivity is calculated for each of the objective function, the value of the constraint condition, and the design variable, and the predicted value of the change amount of the design variable that maximizes the value of the objective function while considering the constraint condition while predicting the tire performance is calculated. The calculation is repeated until the value of the objective function converges (114 to 124 in FIG. 2). The shape of the tire is determined based on the design variables when the value of the objective function converges to the predicted value.

As described above, in the present embodiment, the GL (Global-Local) analysis for predicting the tire performance using a part of the pattern is performed based on the smooth tire model. Therefore, the following three advantages are obtained. Can be obtained. 1: Reduction of calculation time. The present inventor has confirmed that the calculation time required for about one month can be reduced to about two days when the analysis is performed using a fine mesh and a pattern model around the entire circumference. 2: Various models can be easily created. In particular, it is not necessary to prepare the pattern all around in the tire model. 3: It is possible to easily analyze only the pattern of the step portion, the contact portion, and the kick portion.

[0261]

As described above, according to the present invention, a design for giving an optimum value of an objective function satisfying the constraint condition while predicting the performance of the tire under the environment of actual use such as tilting the tire. Variables can be obtained, and a tire can be designed from the design variables, so that the tire development efficiency can be improved and a tire having good performance can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic diagram of a personal computer used in an embodiment of the present invention.

FIG. 2 is a flowchart illustrating a processing flow according to the first embodiment of this invention.

FIG. 3 is a flowchart illustrating a flow of an initial model creation process.

FIG. 4 is a flowchart showing a flow of a tire model creation process.

FIG. 5 is a perspective view showing a tire radial section model.

FIGS. 6A and 6B are image diagrams for explaining the elements at the time of modeling; FIG. 6A is an image diagram for explaining the handling of a rubber portion, and FIG. 6B is an image diagram for explaining the handling of a reinforcing material;

FIG. 7 is a perspective view showing an image in which a tire is modeled.

FIG. 8 is an explanatory diagram for explaining a slip angle and a camber angle.

FIG. 9 is a diagram showing a tire basic model.

FIG. 10 is a perspective view showing a road surface model.

FIG. 11 is a flowchart illustrating a flow of an initial condition setting process.

12A and 12B are explanatory diagrams of approximating a tire model, wherein FIG. 12A is a cross-sectional view of a tire, and FIG. 12B is an explanatory diagram of an approximation process.

FIG. 13 is a flowchart illustrating a flow of a process for determining a design variable.

FIG. 14 is a flowchart showing the flow of Lagrange interpolation processing.

15A and 15B show a tire model for explaining Lagrange interpolation processing, in which FIG. 15A shows a model diagram, and FIG. 15B shows a state in which a virtual line passing through a reference point P is drawn for each dθ in a tire basic model. diagram showing a diagram showing a (C) is a node that has been selected, the relationship such as the distance r _i and yaw angle theta _i of this node.

FIG. 16 is a flowchart showing a flow of a tire performance prediction process.

FIG. 17 is a flowchart illustrating a flow of a process of adding an assignment condition.

FIG. 18 is a diagram showing a tire outer shape.

FIG. 19 is a diagram showing performance evaluation results of a tire.

FIG. 20 is a flowchart illustrating a flow of a process according to the second embodiment of this invention.

FIG. 21 is a flowchart illustrating a flow of a crossover process.

22A and 22B are diagrams illustrating a mountain-shaped mapping function, FIG. 22A is a diagram illustrating a continuous mountain-shaped mapping function, and FIG. 22B is a diagram illustrating a linear mountain-shaped mapping function.

23A and 23B are diagrams illustrating a valley mapping function, FIG. 23A is a diagram illustrating a continuous valley mapping function, and FIG. 23B is a diagram illustrating a linear valley mapping function.

FIG. 24 is a flowchart showing the flow of a mutation process.

FIG. 25 is a schematic configuration diagram of an optimization device according to a third embodiment of the present invention.

FIG. 26 is a schematic block diagram for each function of the optimization device.

FIG. 27 is a conceptual configuration diagram of a neural network.

FIG. 28 is a flowchart showing a flow of a learning process of the neural network.

FIG. 29 is a flowchart showing a flow of operation of the optimization device of the third embodiment.

FIG. 30 is a flowchart illustrating a flow of an optimization process according to the third embodiment.

FIG. 31 is a flowchart showing a processing flow of a pneumatic tire design program according to the fourth embodiment of the present invention.

FIG. 32 is a perspective view showing a basic shape model.

FIG. 33 is a conceptual perspective view showing an input given to a basic shape model.

FIG. 34 is a diagram illustrating a shape of a block near a ground contact edge for explaining a curve fitting process;

FIG. 35 is a diagram showing a block shape as a result of calculation.

FIG. 36 is a perspective view seen from the direction of arrow B in FIG. 35;

FIG. 37 is a diagram illustrating a block shape obtained by performing a curve fitting process on the block shape;

FIG. 38 is a flowchart showing the flow of processing in the case where the shape is approximated for each calculation.

FIG. 39 is a flowchart showing a flow of processing for applying an optimum shape itself to tire design without using shape approximation.

FIG. 40 is a conceptual diagram for explaining a state of wear in a conventional tire shape.

FIG. 41 is a conceptual diagram for explaining a state of wear of a tire shape according to the present invention.

[Explanation of symbols]

 Reference Signs List 10 keyboard 12 computer body 14 CRT 30 tire model FD floppy disk (recording medium) FDU floppy disk unit

Claims (21)

[Claims] 1. A method for designing a pneumatic tire, comprising:
(A) a tire model including at least a tire cross-sectional shape including an internal structure and capable of giving deformation by grounding, a road surface model in contact with the tire model, and at least one of the tire model and the road surface model; Initial conditions related to rolling, an objective function representing a physical quantity for tire performance evaluation, design variables for determining the tire cross-sectional shape or tire structure or pattern shape, tire cross-sectional shape, tire structure, pattern shape, physical quantity for performance evaluation And a constraint restricting at least one of the tire dimensions;
(B) predicting tire performance based on physical quantities generated in the tire model due to the deformation of the tire model; and (c) determining design variables that give the optimal value of the objective function in consideration of the predicted tire performance and constraints. A step of obtaining a value, and (d) a step of designing a tire based on a design variable giving an optimum value of an objective function. 2. In the step (a), at least one of an angular velocity at which the tire model rolls and a moving velocity at which the road surface model moves is determined as an initial condition. 2. The method for designing a pneumatic tire according to 1. 3. The method according to claim 1, wherein, in the step (a), a calculation is performed at the time of filling the internal pressure and at the time of calculating the load, and a tire model to which a rotational displacement or a speed or a linear displacement or a speed is given is determined. 3. The method for designing a pneumatic tire according to 2. 4. The step (b) includes: (e) a step of performing a deformation calculation of the tire model; (f) a step of giving an application condition to the tire model; and (g) a deformation in the step (e). Applying the conditions of step (f) to the tire model after the calculation and allowing the tire model to calculate until a predetermined time has elapsed; (h) obtaining a physical quantity generated in the tire model; and (i) tire performance based on the physical quantity. The method for designing a pneumatic tire according to any one of claims 1 to 3, further comprising: 5. The pneumatic tire designing method according to claim 1, wherein the tire model has a pattern partially. 6. The road surface model according to claim 1, wherein the road surface model determines a road surface condition by selecting a friction coefficient μ representing at least one road surface condition of DRY, WET, on ice, on snow, and unpaved. A method for designing a pneumatic tire according to claim 5. 7. The tire according to claim 1, wherein at least one of a contact area and a contact pressure of a tire model is used as the physical quantity, and a tire WET performance is predicted as the tire performance. 3. The method for designing a pneumatic tire according to 1. 8. The on-ice and snow performance of the tire model is predicted as at least one of a contact area, a contact pressure, and a shearing force on at least one of an ice road surface and a snow road surface of the tire model as the physical quantity. The pneumatic tire designing method according to any one of claims 1 to 7, wherein: 9. In step (c), the sensitivity of the objective function, which is the ratio of the change amount of the objective function to the unit change amount of the design variable, and the constraint, which is the ratio of the change amount of the constraint condition to the unit change amount of the design variable. Based on the sensitivity of the condition, the change amount of the design variable that gives the optimal value of the objective function is predicted while considering the constraint condition, and the value of the objective function and the design variable when the design variable is changed by an amount corresponding to the predicted amount Calculates the value of the constraint condition when the value of the target variable is changed by an amount corresponding to the predicted amount, and calculates the value of the design variable that gives the optimal value of the objective function based on the predicted value and the calculated value while considering the constraint condition. Item 1
How to design pneumatic tires. 10. A function representing the shape of at least one of a carcass line, a folded ply line, a line representing a tire outer shape, a line representing a tire crown shape, and a reinforcing material line, Gauge distribution, rubber chafer gauge distribution, side rubber gauge distribution, tread rubber gauge distribution, tread base rubber gauge distribution, inner surface reinforcing rubber gauge distribution, belt-to-belt rubber gauge distribution, and belt end rubber gauge distribution. Variables representing the gauge distribution of at least one tire rubber member; variables representing the structure of at least one belt portion of the angle, width, cord type, and driving density of each belt layer; block shapes and sipe positions; A variable representing the shape of at least one pattern of length and length; The pneumatic tire designing method according to claim 1, characterized in that it comprises Kutomo one. 11. In the step (a), a selection target group including a plurality of tire models including at least a tire cross-sectional shape including an internal structure and capable of being deformed by grounding is determined, and the selection target group is selected. For each tire model, an objective function representing a physical quantity for tire performance evaluation,
From design variables for determining a tire cross-sectional shape or a tire structure or a pattern shape, a constraint condition for restricting at least one of a tire cross-sectional shape, a tire structure, a pattern shape, a physical quantity for performance evaluation and a tire dimension, and an objective function and a constraint condition An adaptation function that can be evaluated is determined, and a road surface model that comes into contact with the tire model and an initial condition related to rolling of at least one of the tire model and the road surface model are determined. Selecting two tire models from the selection target group based on the function, crossing design variables of the respective tire models with a predetermined probability to generate a new tire model, and selecting one of the design variables of at least one of the tire models. Change the design variables by at least one of the following: The objective function, the constraint condition and the adaptive function of the tire model obtained are obtained, the tire model and the tire model in which the design variables are not changed are stored, and the stored tire model is repeated until the stored number of tire models reaches a predetermined number. Determine whether the new group consisting of tire models satisfies a predetermined convergence condition, and when the convergence condition is not satisfied, repeat the new group as the selection target group until the selection target group satisfies the predetermined convergence condition, 2. The air according to claim 1, wherein a value of a design variable that gives an optimum value of the objective function is determined from the predetermined number of tire models stored when the predetermined convergence condition is satisfied, while considering the constraint condition. How to design a tire with a hole. 12. In the step (c), for the tire model in which the design variables are changed, the sensitivity of the objective function, which is the ratio of the change amount of the objective function to the unit change amount of the design variables, and the change in the unit amount of the design variables. Based on the sensitivity of the constraint, which is the rate of change of the constraint, the amount of change of the design variable that gives the optimal value of the objective function is predicted while considering the constraint, and the design variable is changed by an amount corresponding to the predicted amount. The value of the objective function and the value of the constraint when the design variable is changed by an amount corresponding to the predicted amount are calculated, and the adaptive function is obtained from the value of the objective function and the value of the constraint to obtain the tire model and the design. 12. The pneumatic tire designing method according to claim 11, wherein a tire model in which a variable is not changed is stored and the stored tire basic model is repeated until a predetermined number of tire models are stored. Law. 13. In the step (a), a non-linear correspondence between design parameters of a tire model including at least a tire cross-sectional shape including an internal structure and capable of giving deformation by grounding and performance of the tire is related. A conversion system is defined, and a constraint condition for limiting at least one of the tire performance and the manufacturing condition of the tire is defined as the constraint condition. In the step (c), the conversion system defined in the step (a) is used. Determining a tire design parameter that gives an optimal value of the objective function based on the objective function and the constraint condition, and designing the tire based on the tire design parameter in the step (d). 2. The method for designing a pneumatic tire according to 1. 14. In the step (c), a design parameter of the tire model is determined as a design variable, and an optimum value of an objective function is given by using a conversion system determined in the step (a) while considering a constraint condition. 2. The tire according to claim 1, wherein a value of a design variable is determined, and in the step (d), the tire is designed based on a design variable that gives an optimum value of the objective function.
3. The method for designing a pneumatic tire according to 3. 15. In the step (a), the conversion system is constituted by data of a multilayer feed-forward type neural network learned so as to convert design parameters of the tire model into the tire performance. The method for designing a pneumatic tire according to claim 13 or 14. 16. In the step (a), the shape of a single block including the internal structure, the pattern shape of a part of the tire crown portion including the internal structure, and the land extending in the circumferential direction of the tire including the internal structure. 1 selected from the shape of the part
In addition to defining a basic shape model that represents two shapes,
The shape basic model is given at least one input condition, and a tread shape representing at least a part of the shape of the block alone or the shape of the pattern or the shape of the land portion is used as a design variable, and the tire contact pressure under the input condition is calculated. 3. The method according to claim 1, wherein
3. The method for designing a pneumatic tire according to 1. 17. The constraint condition further defines at least one of a tire contact area and a change range of a design variable. In the step (c), an optimum value of the objective function is given in consideration of the constraint condition. The method of designing a pneumatic tire according to claim 16, wherein the value of the design variable is changed up to the value. 18. The pneumatic tire design method according to claim 17, wherein the design variable changes a design variable of at least one of a portion higher and a portion lower than the tire average contact pressure. 19. A storage unit storing a program for executing the processing of steps (a) and (b) according to claim 1 to 18, reading said program from said storage unit, and Prediction means for predicting the performance of the tire by a physical quantity generated in the tire by deformation using the design parameters; conversion system calculation means for obtaining a non-linear correspondence between the design parameters of the tire and the performance of the tire; the tire performance And an input function for inputting as an optimization item, defining a constraint condition for restricting an allowable range of at least one of the tire performance and the tire manufacturing condition, and using the conversion system calculation means. A tire design parameter that gives an optimal value of an objective function is determined based on the optimization item input by the input means. And an optimization calculation means. 20. The optimization according to claim 19, wherein said conversion system calculation means is a multilayer feedforward neural network learned to convert the tire design parameters into the tire performance. Analysis device. 21. A storage medium storing a tire optimization analysis program for designing a tire by a computer, the optimization analysis program comprising: steps (a) and (b) according to claim 1 to claim 18. A) predicting the performance of the tire by a physical quantity generated in the tire by deformation using the design parameters of the tire, reading the prediction program, and calculating the tire design parameters and the tire A non-linear correspondence relationship with performance is determined, an objective function representing the performance of the tire is determined, and a constraint condition that restricts at least one allowable range of the performance of the tire and manufacturing conditions of the tire is determined. Tire design that gives the optimal value of the objective function based on the relationship, the objective function and the constraint The storage medium storing the optimization analysis program of the tire, characterized by designing the tire based on the design parameters of the tire in search of parameters.