JP2009078757A - Road surface displacement inferring device for vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a road surface displacement inferring device capable of reducing amount of required computation and obtaining sufficient responsiveness by utilizing an existing sensor in an active suspension system of a vehicle effectively. <P>SOLUTION: In this road surface displacement inferring device, the active suspension system uses the rule for controlling an actuator and minimizing the integration of function having at least the term expressing the energy to be transmitted to a vehicle body and a wheel to be controlled from the actuator, the term expressing vibration conditions of the vehicle body and the wheel, and the term expressing balance of total energy of the vehicle body and the wheel. In this case, since adverse affecting due to error in inferring road surface displacement is slight, the expression for inferring it is substituted for the following simple expression (A): (A)q<SB>0</SB>=α×M<SB>1</SB>×q<SB>1</SB>"/K<SB>1</SB>(wherein, α denotes correction coefficient, M<SB>1</SB>denotes mass of an unsprung member 3, q<SB>1</SB>" denotes acceleration of the unsprung member 3, and K<SB>1</SB>denotes spring constant of the unsprung member 3). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an active suspension system in which at least a vertical control force is applied to a wheel by an actuator, and more particularly to an apparatus for estimating a displacement state such as unevenness on a road surface on which a vehicle travels.

  Conventionally, as an active suspension system of this type, for example, an electromagnetic suspension device disclosed in Patent Document 1 is suspended by an actuator (linear motor, ball screw mechanism, etc.) disposed between a vehicle body and a wheel. It is known to make the stroke.

Further, in the active stabilizer device described in Patent Document 2, a torsion bar type stabilizer is divided into left and right parts, and they are driven in opposite directions by actuators provided in the middle so that the left and right wheels are reversed in phase. Thus, it is possible to variably control the roll rigidity of the vehicle, such as suppressing the bounce and rebound of the left and right wheels.
Japanese Patent Laid-Open No. 2007-083813 JP 2006-232023 A

  By the way, in the active suspension system as described above, it is sometimes necessary to measure or estimate the unevenness of the road surface in order to control the actuator, but only for that purpose a sensor (for example, an ultrasonic sensor) Providing is connected to the cost increase and is not preferable.

  On the other hand, when estimating the road surface displacement based on the behavior of wheels detected by existing sensors in the system, the amount of calculation generally increases considerably, resulting in a large response delay and a high calculation load on the CPU. There is a problem because it becomes too much.

  A method for estimating road surface unevenness using a Kalman filter has also been proposed, but this method estimates the current road surface unevenness based on past driving data, and requires a certain level of responsiveness. It cannot be used for control.

  SUMMARY OF THE INVENTION In view of these points, an object of the present invention is to provide a road surface displacement estimating apparatus that requires a small amount of calculation and can obtain sufficient responsiveness while utilizing an existing sensor in a vehicle active suspension system. There is.

  The inventor of the present invention has obtained a new finding that when a certain type of optimal control law is used as the control law of the actuator of the active suspension system, the adverse effect due to the estimation error of the road surface displacement becomes minor. Thus, the above-mentioned object is achieved by replacing the estimated calculation formula for this purpose with a simple formula.

  That is, the invention according to claim 1 is a vehicle road surface displacement that estimates a displacement of a road surface on which the vehicle travels in an active suspension system of a vehicle in which at least a vertical control force is applied to a wheel by an actuator. The estimation device is the target.

And the control law for calculating the control amount to the actuator is at least a term representing the energy transmitted from the actuator to the vehicle body and the wheel to be controlled, and a term representing the vibration state of the vehicle body and the wheel, When the optimal control law is such that the integral of the function having the term representing the total energy balance of the vehicle body and wheels is minimized,
The road surface displacement estimation device includes acceleration detection means for detecting vertical acceleration of an unsprung member of a suspension including a wheel, at least acceleration of the unsprung member detected by the acceleration detection means, mass of the unsprung member, And road surface displacement estimating means for estimating the road surface displacement based on the spring constant of the unsprung member.

  In the above-described configuration, the mass of the unsprung member such as a tire and a wheel can be set in advance, and the spring constant of the tire can be obtained relatively easily. If the acceleration is detected, it is possible to estimate the road surface displacement with a relatively small calculation amount by a simple estimation calculation without performing a complicated calculation.

For example, based on the acceleration q ₁ ″ of the unsprung member, the mass M _{1 of} the unsprung member, and the spring constant K ₁ of the unsprung member, the road surface displacement q ₀ is estimated and calculated by the following (formula A). In (Equation A), α is a correction coefficient, and it has been experimentally confirmed that a relatively high estimation accuracy can be obtained by setting the value to about 4-6. Yes.

q ₀ = α · M ₁ · q ₁ ″ / K ₁ ... (Formula A)
The above (Formula A) is a simplified formula for calculating the road surface displacement based on the behavior of the vehicle body and wheels that can be detected by an acceleration sensor, a displacement sensor or the like that is usually provided in an active suspension system. Other than the acceleration q ₁ ″ of the unsprung member, it can be regarded as substantially constant regardless of the road surface condition, and therefore the calculation amount is very small.

  On the other hand, it is inevitable that the estimation error of the road surface displacement by such a simple formula becomes relatively large, but the control of the active suspension system performed using this estimation result follows the optimal control law as described above. If it is, the adverse effect due to the estimation error of the road surface displacement will be negligible, so there will be no practical problem.

The mass M _{1 of} the unsprung member, the spring constant K ₁ , and the correction coefficient α in the (formula A) can be stored in advance in the storage means for each vehicle type (Claim 3). Because it is only necessary to make input settings for, it is highly convenient. Note that the spring constant K ₁ of the unsprung member, that is, the spring constant of the tire changes depending on the air pressure, and therefore it is desirable to correct it according to the output from the air pressure sensor attached to the tire.

  Further, the value of the correction coefficient α may be changed according to the traveling state of the vehicle (claim 4). That is, (Formula A) shows that the vertical acceleration generated in the unsprung member due to the road surface unevenness is approximately proportional to the road surface unevenness. The magnitude of the unsprung vertical acceleration caused by is also affected by, for example, the vehicle speed. If the value of the correction coefficient α is changed in accordance with, for example, the vehicle speed in consideration of this point, the influence is taken into account Estimation can be performed.

  As described above, according to the road surface displacement estimating apparatus for a vehicle according to the present invention, when the actuator in the active suspension system is controlled according to a certain optimal control law, the adverse effect due to the road surface displacement estimation error is adversely affected. Focusing on the fact that the calculation is small, the calculation formula for the estimation calculation is reduced while simplifying the estimation calculation formula as in the above (formula A), and sufficient Control responsiveness can be obtained.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. It should be noted that the following description of the preferred embodiment is merely illustrative in nature, and is not intended to limit the present invention, its application, or its use.

(Schematic configuration of active suspension system)
In Figure 1, an automobile A equipped with an active suspension system S according to the present invention (vehicle) schematically shows, in this example, as shown in FIG. (A), the suspension 1 _FR of the front and rear left and right 4 wheels , 1 _FL , 1 _RR , 1 _RL are provided with electromagnetic actuators 2, 2,. The suspensions 1 _FR 1, 1 _FL ,... Of each wheel include a so-called unsprung member 3 including a tire 3 a, a wheel 3 b, and a suspension member (not shown) such as a wheel support that supports the tire 3 a, a coil spring 4 (plate Springs, torsion bars, air springs, etc.) and shock absorbers 5 are connected to the vehicle body B, and each wheel is connected by an electromagnetic actuator 2 provided between the vehicle body B in parallel with the coil springs 4 and the like. At least a control force in the vertical direction is applied to the.

As shown schematically in FIG. 2B, the suspension 1 is mechanically composed of an unsprung member 3 including a tire 3a and a wheel 3b, and a sprung spring connected to the unsprung member 3 by a coil spring 4 and a shock absorber 5. It can be regarded as a two-degree-of-freedom vibration system composed of the member 6 (mainly the mass shared by the vehicle body B). In this case, as shown in the figure, the mass of the unsprung member 3 is M ₁ , its spring constant is K ₁ , the spring constant of the coil spring 4 is K ₂ , the damping coefficient of the shock absorber 5 is C, and mass is referred to as _{M 2.}

Further, the unevenness of the road surface to be grounded of the tire 3 a, that is, the vertical displacement is q ₀ , the vertical displacement of the unsprung member 3 is q ₁ , and the vertical displacement of the sprung member 6 is q _2, and is input to the electromagnetic actuator 2. The controlled variable u is represented by u, and the force generated by the electromagnetic actuator 2 driven thereby is represented by f _a (not shown). Note that generation force f _a of the electromagnetic actuator 2, a direction in which push apart the unsprung member 3 and sprung 6 together with a positive value, the direction to draw both the negative value.

  As the electromagnetic actuator 2, a linear motor is used as an example. A rod connected to the unsprung member 3 has a permanent magnet, and a sprung member 6 side has a drive coil on its side to surround it. Has been placed. The forward / backward driving force of the rod is controlled by power supply control to the driving coil, and the control force is applied to the unsprung member 3 and the unsprung member 6, respectively. Of course, you may connect a rod on a spring.

And the drive control of the electromagnetic actuators 2, 2,... For each suspension 1 _FR , 1 _FL ,. As shown schematically in FIG. 2, the vehicle body B of the automobile A detects the acceleration q ₂ ″ in the vertical direction corresponding to the attachment site (on the spring) of the suspension 1 _FR , 1 _FL,. , And the stroke sensors 12, 12,... That detect the stroke q _s (= q ₂ −q ₁ ) of the suspension 1, and for detecting a predetermined vehicle state quantity. The vehicle state detection sensor 13 is also provided.

The controller 10 receives the signals from the sensors 11 to 13 and controls the electromagnetic actuators 2, 2,... For each suspension 1 _FR , 1 _FL ,. Actively change. Specifically, it absorbs input due to road surface irregularities and the like, reduces vibration transmission to the vehicle body, and suppresses changes in the posture of the vehicle body due to inertial forces, thereby achieving both a high level of riding comfort and exercise performance.

More specifically, the controller 10 includes a vertical displacement of the unsprung member 3 for each suspension 1 _FR , 1 _FL ,... Based on signals from the acceleration sensors 11, 11,. Suspension state quantity detection for calculating q ₁ , its speed q ₁ ′ and acceleration q ₁ ″, and the vertical displacement q ₂ and its speed q ₂ ′ of the sprung member 6, that is, the suspension state quantity representing the operating state of the suspension 1. And a traveling state detection unit 10b that receives various signals from the vehicle state detection sensor 13 and detects various traveling states such as sudden acceleration, sudden braking, or sudden steering, for example. ing.

In addition, the controller 10, on the basis of the suspension state quantity detecting unit acceleration q ₁ unsprung member 3 which is calculated by 10a _"or the like, the road surface displacement estimation unit 10c for estimating the displacement q ₀ of the road running of the automobile A ( .. Based on the road surface displacement estimating means), the road surface displacement q ₀ , the suspension state quantities (q ₁ , q ₁ ′, q ₁ ″, q ₂ , q ₂ ′), etc. A control amount calculation unit 10d for calculating the control output u, and a correction control unit 10e (correction) for correcting the calculation method in the road surface displacement estimation unit 10c and the control amount calculation unit 10d according to the traveling state of the vehicle. Means) is also provided.

  The functions of the suspension state quantity detection unit 10a, the traveling state detection unit 10b, the road surface displacement estimation unit 10c, the control amount calculation unit 10d, and the correction control unit 10e are executed by a predetermined program executed by the CPU of the controller 10. In this sense, the controller 10 includes the units 10a to 10e in the form of a software program.

In particular, the control amount calculation unit 10d positively strokes the suspensions 1 _FR , 1 _FL ,... By operating the electromagnetic actuators 2, 2,. An arithmetic expression for the control output u is set to control the electromagnetic actuators 2, 2,... So as to suppress vibration of the vehicle body B and minimize energy consumption therefor. In other words, the arithmetic expression of the control output u is in accordance with an optimal control law as described in detail below.

(How to find the optimal control law)
Next, regarding the arithmetic expression of the control output u set in the control amount arithmetic unit 10d of the controller 10 as described above, in particular, a technique for deriving the arithmetic expression for the optimal control as described above, that is, the optimal control problem The method of solving will be described in detail.

-Basic concept-
First, the basic concept will be explained. In general, in the optimal control problem, the characteristics of the controlled object are described by an equation of motion, and a control law is determined that maximizes or minimizes the evaluation function defined from various viewpoints for the system that controls this. Usually, it is not easy to derive such a control law analytically.

  In this regard, the inventors of the present application, as a method for analytically solving a relatively wide range of optimal control problems including nonlinear systems of mechanical dynamics systems, pay attention to the energy balance of the control target and control rules that asymptotically stabilize the system. We have devised a method that can easily guide you.

  In this method, the characteristic of the controlled object is not described by the equation of motion, but the balance equation (1) of the total input / output power of the controlled object is used as follows. This equation (1) is a vector display of the equation of motion for each degree of freedom of the system, and this is multiplied by the velocity vector. The input power includes not only control input but also disturbance input. Since the control target is not limited to passive elements, there is an energy source inside, and if this affects the movement, it is treated as a disturbance input.

In the above equation (1), d, e, q, u, v, zεR ⁿ , MεR ^{n × n} are positive definite inertia matrices, and n is the degree of freedom of the controlled object. q is a generalized coordinate, u is a control input, and the number of independent actuators is n. ν is a force input disturbance, and z is a displacement input disturbance. u and ν act directly on inertia, and z acts on inertia via unsprung. d is Coriolis force, centrifugal force, damping force, etc., e is potential force.

  Even if the object to be controlled is non-linear, if the control device is rationally designed, u can be directly applied to M as shown in equation (1). Assuming a rationally designed mechanical mechanical system, consider the following evaluation function J for this system.

In the equation (2), g is a scalar function that gives an evaluation of control performance, and u ^T q ′ is the power applied from the actuator of the control device to the controlled object, that is, the energy transmitted from the actuator. r is a weight coefficient and is a positive definite value. This method is intended for real-time control and assumes a finite evaluation interval. Finding the control amount u (t) that minimizes the equation (2) is the optimal control problem.

  First, in order to obtain the necessary conditions for optimal control, the following scalar function L is defined.

In the formula (3), κ is an undetermined constant. The inside of {} on the right side is the same as the left side of the equation (1) and is the total power balance of the controlled object, so it satisfies the energy conservation law and is always zero. Therefore, the condition for minimizing the integral (functional) of the function L represented by the equation (3) gives the condition for minimizing the equation (2).

  Therefore, the following equation (4) in which the variational principle with q as a variable is applied to the integral of L gives a necessary condition for optimal control with respect to the control input u. Since L is a linear expression with respect to u, ∂L / ∂u has no meaning, and the following equation (4) includes all information related to control.

However, since L includes two derivatives of q, the third term is added to the general Euler equation.

  When the equation (4) is integrated and the integration constant is zero, the following equation (5) is obtained. Substituting the above equation (3) into this equation (5) and writing the first to third terms on the left side as the first to third rows in order, the following equation (6) is obtained.

  Then, the control law is obtained from the above equation (6) as in the following equation (7).

  Thus, since the relationship between q and u that minimizes the evaluation function J can be directly derived, the process of solving the two-point boundary value problem using the principle of optimality as in the conventional general method is unnecessary. Become. κ is an undetermined constant, but when κ = 0, u is determined only by the parameters of the evaluation function, whereas when κ = ∞, u is determined only by the parameter to be controlled. , For u to be optimal, must be a non-zero finite value.

  The first line of the equation (7) is control for the external force ν and q dependency of inertia, the second line is control for Coriolis force, centrifugal force and damping force, and the third line is potential force and its q dependency and external force z. The fourth line is a control for reducing the evaluation function. Equation (7) includes an unexecuted calculus term. However, if all external forces and state quantities can be detected or estimated, these can be executed in real time.

  Although the integral constant is set to zero in the above formula (5), it can be seen from the above results that the integral constant gives a constant bias to the control law, so that it is appropriate to set it to zero. The same applies to the integral in equation (7).

  Further, in the equation (7), it is assumed that the number of actuators and the degree of freedom of the system are the same, and the following processing is necessary when the number of actuators is small. For example, when the actuator is placed between two independent inertias, the constraint condition is included in the control vector, and the optimal control law is obtained by weighting and adding the two control laws. Good.

That is, if u _i and u _{i + 1} are optimal control rules when they are derived as independent controls, and ρ _i and ρ _{i + 1} are weighting factors, then control output u _opt = ρ _i u _i + ρ _{i + 1} u _{i + 1} and Become. Note that the values of the weight coefficients ρ _i and ρ _{i + 1} are not theoretically derived, but depend on the structural features of the controlled object.

-Suspension system-
In accordance with the above concept, an arithmetic expression of the control output u of the electromagnetic actuator 2 in the active suspension system S of this embodiment, that is, the optimum control law u is derived. First, in the suspension system represented by the model of FIG. 2 described above, the evaluation function J of Equation (2) is described using the variable q as shown in Equation (8) below.

In the equation (8), r ₁ and r are weighting factors, and the weight r ₁ is a value (for example, 5 × 10 ⁻³⁾ such that the vibration suppression of the unsprung mass is substantially the same as that of a conventional automobile. ) And the weight r is set to a value (for example, 10 × 10 ⁻⁵ ) such that the evaluation of the vibration state and the evaluation of the energy are approximately the same.

  Further, the function L of the equation (3) is described as the following equation (9).

Substituting the equation (9) into the equation (5) yields the following equation (10) for q ₁ , thereby obtaining the control law of the following equation (11).

Similarly, the following equation (12) is obtained for q ₂ , thereby obtaining the control law of equation (13).

  In summary, the final control law is as follows.

u _opt = ρ ₁ u ₁ + ρ ₂ u ₂ (14)
Here, assuming that the sum of the values of the weight coefficients ρ ₁ and ρ ₂ is ρ ₁ + ρ ₂ = 2 and the sensitivity of the control output u to the inertia M ₁ and M ₂ considering the weight of the evaluation function is equal, Thus, the optimal control output u = u _opt can be obtained as the following expression (16).

In Equation (16), ρ ₁ = 2M ₁ / (M ₁ + r ₁ M ₂ ) and ρ ₂ = 2r ₁ M ₂ / (M ₁ + rM ₂ ).

  The control law derivation by this method is up to the above equation (16), which gives the necessary conditions for optimal control. For the undetermined constant κ in equation (16), a value that minimizes the evaluation function J is determined by an exploratory method. For example, the value of κ that minimizes the evaluation function J can be obtained by simulation of random road surface input.

The arithmetic expression (16) derived as described above represents the function L of the expressions (3) and (9), that is, the energy transmitted from the electromagnetic actuator 2 to the controlled object (sprung member 6 and unsprung member 3). This is in accordance with an optimal control law that minimizes the integral of the function L having terms, terms representing their vibration states, and terms representing the total energy balance of the controlled object. By controlling the electromagnetic actuator 2 with the control output u obtained in 16), the suspensions 1 _FR , 1 _FL ,... _Are optimally operated while suppressing energy consumption, and vibrations of the wheels and the vehicle body B are effectively suppressed. be able to.

(Estimation of road displacement)
Next, estimation of the road surface displacement q ₀ in the road surface displacement estimation unit 10c of the controller 10 will be described as a feature of the present invention. As described above, in the system S of this embodiment, the road surface displacement q ₀ is used in the calculation of the control output u to the electromagnetic actuators 2, 2,... According to the equation (16), which is generally an ultrasonic sensor. However, the addition of a new sensor leads to an increase in the cost of the system, which is not preferable.

Therefore, in this embodiment, among the suspension state quantities (q ₁ , q ₁ ′, q ₁ ″, q ₂ , q ₂ ′) detected by the sensors 11 to 13 existing in the system S, mainly the unsprung member 3. On the basis of the acceleration q ₁ ″, an operation for estimating the road surface displacement q ₀ from the behavior of the wheels and the vehicle body B by the input from the road surface is performed.

That is, first, the behavior of the wheel by the ground forces f _g each of the suspension 1 _{FR, 1} FL, _... for each is expressed by the following equation (17), whereas, the behavior of the spring force f _sus and the wheel of the coil spring 4 Is expressed by the following equation (18). However, the tire 3a has a damping coefficient C _tire .

From the equations (17) and (18), the ground contact force _fg is expressed by the following equation (19). Note that q _s is a suspension stroke detected by the stroke sensor 12 (q _s = q ₂ −q ₁ ).

If the equation (19) is substituted into the equation (17) and solved as an ordinary differential equation, an arithmetic expression for estimating the road surface displacement q ₀ can be obtained. However, the arithmetic expression thus obtained becomes complicated and the amount of calculation because becomes many, on which becomes high calculation load of the CPU is the estimation calculation of the road surface displacement q ₀ by this, there is a concern that calculation time is long control response delay of the electromagnetic actuator 2 becomes large. Moreover, since the equation (17) includes a large variation value like the damping coefficient C _tire of the tire 3a, it cannot be said that it can be estimated with high accuracy.

Therefore, the second term on the left side from Equation (17), the fourth term on the right side from Equation (19), and so on are simplified by deleting the damping force term, and then the equation (19 ) Is substituted into equation (17), the following equation (20) is obtained, and when this is arranged for q ₀ , equation (21) is obtained.

Equation (21) is omitted the effects of attenuation, which primarily to estimate the road surface displacement q ₀ from the acceleration of the unsprung member 3, if to estimate the road surface displacement q ₀ Using this arithmetic expression The calculation amount is considerably reduced. Further, if the first term (the term of the unsprung acceleration q ₁ ″) in the right side {} of the equation (21) is deleted and multiplied by the correction coefficient α, q ₀ = αM ₁ q ₁ ″ / K ₁ ... (22)

This equation (22) is to estimate the displacement state of the road surface substantially based only on the acceleration q ₁ ″ of the unsprung member 3, and since the amount of calculation is very small, the road surface displacement q ₀ estimated in this way is obtained. becomes very advantageous in increasing the responsiveness of the suspension control based. on the other hand, although such the estimation error of the road surface displacement q ₀ According to simple calculation expression becomes large, by using the estimated value above as a result of such suspension control of adverse effects due to an estimation error of the road surface displacement q ₀ is immaterial, it has been found that practically no problem.

FIG. 3 shows the same estimation as in the case where the optimum control of the present invention is performed based on the control output u of the equation (16) based on the road surface displacement q ₀ estimated using the equations (21) and (22). The results of control are compared by simulation in the case where LQ control, which is conventionally known optimum control, is performed using values. FIG. 6A shows the result of comparing the control results with the above-described evaluation function J, and FIG. 6B shows the result of comparison with the general evaluation function J _LQ in the LQ control.

From the figure (a), if the estimated value of the road surface displacement q ₀ of the equation (21) is used, the evaluation function J becomes a very small value in both the optimum control of the present invention and the conventional LQ control. It can be seen that the same optimal control can be performed. On the other hand, when the estimated value of the road surface displacement q ₀ in Expression (22) is used, the evaluation function J becomes very small in the optimum control of the present invention, but the value of the evaluation function J becomes considerably large in the LQ control. . From this, the estimated value of equation (22) includes an error that cannot be ignored, but when performing the optimal control of the present invention, the adverse effect due to this is negligible, and it can be seen that there is virtually no problem. .

A similar tendency can be read from FIG. 5B, and if the estimated value of the equation (21) is used, the value of the evaluation function J _LQ is almost the same between the optimum control and the LQ control of the present invention, It can be seen that the value of the evaluation function J _LQ becomes considerably large only in the case of the LQ control when the estimated value of the equation (22) is used. Note that when the estimated value of equation (21) is used, the value of the evaluation function J _LQ is smaller in the LQ control than in the optimal control of the present invention. This evaluation function J _LQ is optimal for the LQ control. It is because it has become.

Further, FIG. 4 shows a case where the optimum control and the LQ control of the present invention are executed using the estimated value of the road surface displacement q ₀ while changing the value of the correction coefficient α in the equation (22). The evaluation results are shown in comparison. The evaluation of the optimum control of the present invention was performed by the evaluation function J, and the evaluation of the LQ control was performed by the evaluation function J _LQ . From the figure, the evaluation function J _{LQ of the} LQ control becomes a large value regardless of the value of the correction coefficient α, while the value of the evaluation function J according to the control of the present invention is α = 4 to 10, especially 4 It can be seen that the value is very small at ˜6.

-Road surface displacement estimation procedure-
Next, the procedure for estimating the road surface displacement q ₀ in the controller 10 will be specifically described with reference to the flowchart of FIG. First, in step S1 after starting the illustrated flow, signals from the sensors 11 to 13 are input to detect the sprung acceleration q ₂ ″ and the suspension stroke q _s, and in the subsequent step S2, the suspension stroke q _{s is set} to 2. The stroke acceleration q _s ″ is calculated by differential differentiation.

Next, in step S3 by subtracting the "stroke acceleration _{q s} to" the sprung acceleration _{q 2,} "to calculate the _{(q 1"} acceleration _{q 1} unsprung member _{3 = q 2 "-q s"} ) and In step S4, the unsprung acceleration q ₁ ″ thus obtained is substituted into equation (22) to estimate the road surface displacement q ₀ and then the process returns.

Here, the correction coefficient α, the unsprung mass M ₁ , and the unsprung spring constant K ₁ included in the equation (22) are set in advance for each vehicle type by experiment or the like, and the memory or the like (storage means) of the controller 10 is stored. ) Is stored. In particular the spring constant K ₁ under spring, i.e. the spring constant of the tire 3a, since the voice is changed pneumatically, when the air pressure sensor is attached to for example the wheel 3b, the spring constant K in accordance with the output _It is preferable to correct ₁ .

Further, as described above with reference to FIG. 4, the value of the correction coefficient α is set by experiments or the like in the range of α = 4 to 6, but in this embodiment, the correction coefficient α is corrected according to the traveling state of the automobile A. Is done. That is, the correction control unit 10e of the controller 10 performs correction so that the value of the correction coefficient α decreases as the vehicle speed increases, for example. As a result, the road surface displacement q ₀ is estimated by taking into account the influence of the vehicle speed on the unsprung acceleration q ₁ ″.

  The procedure of steps S1 to S3 is performed by the suspension state quantity detection unit 10a of the controller 10, and the procedure of step S4 is performed by the road surface displacement estimation unit 10c. The suspension state quantity detection unit 10a and the road surface displacement estimation unit 10c constitute a vehicle road surface displacement estimation device that estimates the displacement state of the road surface on which the automobile A travels.

Therefore, according to the vehicle road surface displacement estimating apparatus according to this embodiment, the acceleration q ₁ ″ of the unsprung member 3 detected by the existing sensors 11 to 13 in the active suspension system S of the automobile A, and the unsprung member Since the road surface displacement q ₀ is estimated and calculated by the simplified equation (22) based only on the mass M _{1 of} 3 and the spring constant K ₁ , the calculation amount is very small. The response of the suspension control to be used can be made sufficiently high.

Thus, the estimation result of the road surface displacement q ₀ by the simplified calculation formula includes a relatively large error. However, the optimum control of the suspension as in this embodiment, that is, the energy transmitted to the controlled object is expressed as follows. When optimal control is performed so as to minimize the integral of the function L having a term to represent, a term representing the vibration state (damping performance) of the controlled object, and a term representing the total energy balance of the controlled object In this case, the adverse effect due to the estimation error of the road surface displacement q ₀ is slight, and there is no practical problem.

  That is, the existing sensors 11 to 13 are used in the system S, and active suspension control having practically sufficient accuracy and responsiveness is performed without causing an increase in cost. It is possible to improve the riding comfort of the automobile A and to improve the ground contact property of the wheels while suppressing the energy consumption for the operation of the vehicle.

In this embodiment, the mass M ₁ , the spring constant K ₁ , and the correction coefficient α of the unsprung member 3 in the calculation formula (22) for estimating the road surface displacement are stored in advance in the memory of the controller 10 for each vehicle type. Thus, in the production line of the automobile A, it is not necessary to measure the mass M ₁ and the spring constant K ₁ of the unsprung member 3 one by one.

Furthermore, the value of the correction coefficient α in the calculation formula (22) is corrected according to the traveling state of the automobile A such as the vehicle speed, and the road surface displacement q ₀ is estimated in consideration of the influence of the traveling state. be able to.

(Other embodiments)
In addition, the structure of the road surface displacement estimation apparatus which concerns on this invention is not limited to the said embodiment, Various other structures are also included. That is, for example, in the above-described embodiment, the electromagnetic actuators 2, 2,... Are provided in the suspensions 1 _FR , 1 _FL ,. Only the electromagnetic actuators 2 and 2 may be provided.

  In addition to the linear motor exemplified as the actuator, for example, an hydraulic / pneumatic cylinder, a piezoelectric element, or the like can be used, and an electric motor and a ball screw mechanism can be combined to form an actuator.

  Further, the sensors in the active suspension system S are not limited to the acceleration sensor 11 and the suspension stroke sensor 12 of the illustrated vehicle body B, and various acceleration sensors, speed sensors, and displacement sensors can be used in combination.

  Furthermore, the above-described embodiment is only an example for active suspension control, and it is only necessary to follow a control law that minimizes an evaluation function defined from various viewpoints in addition to the illustrated evaluation function J. Such a control law minimizes the integral of a function (L in relation to equations (3) and (9)) obtained by adding a term representing the total energy balance of the controlled object to the term of the evaluation function, as in the above embodiment. It must be safe.

  Furthermore, the road surface displacement estimation apparatus according to the present invention can be applied to vehicles other than automobiles.

  The road surface displacement estimation device according to the present invention can estimate and calculate the road surface displacement with a simplified calculation formula while using an existing sensor in a vehicle active suspension system. It can be obtained and is suitable for mounting on an automobile.

FIG. 2 is a diagram schematically showing an automobile (a) equipped with an active suspension system according to the present invention and a suspension (b). It is a block diagram which shows schematic structure of an active suspension system. It is a graph which compares and shows the evaluation result of the optimal control of this invention, and LQ control about each, switching two estimation arithmetic expressions of a road surface displacement. It is a graph which compares and shows the evaluation result of the optimal control of this invention, and LQ control, changing the value of a correction coefficient. It is a flowchart figure which shows the procedure of the estimation calculation of a road surface displacement.

Explanation of symbols

A car (vehicle)
B Body S Active suspension system 1 Suspension 2 Electromagnetic actuator 3 Unsprung member 3a Tire 3b Wheel 4 Coil spring 5 Shock absorber 6 Sprung member 10 Controller 10a Suspension state quantity detection unit (acceleration detection means)
10b Traveling state detection unit 10c Road surface displacement estimation unit (road surface displacement estimation means)
10d Control amount calculation unit 10e Correction control unit (correction means)
11 Acceleration sensor 12 Stroke sensor

Claims (4)

In a vehicle active suspension system in which at least vertical control force is applied to wheels by an actuator, a road surface displacement estimation device for a vehicle that estimates displacement of a road surface on which the vehicle travels,
The calculation of the control amount to the actuator includes at least a term representing energy transmitted from the actuator to the vehicle body and wheels to be controlled, a term representing the vibration state of the vehicle body and wheels, and the total energy of the vehicle body and wheels. Which is performed according to an optimal control law that minimizes the integral of a function having a term representing a balance,
Acceleration detecting means for detecting the vertical acceleration of the unsprung member of the suspension including the wheel;
Road surface displacement estimating means for estimating road surface displacement based on at least the acceleration of the unsprung member detected by the acceleration detecting means, the mass of the unsprung member, and the spring constant of the unsprung member. The road surface displacement estimation apparatus for vehicles. The road surface displacement estimating means calculates the road surface displacement q ₀ by the following (formula A) based on the acceleration q ₁ ″ of the unsprung member, the mass M _{1 of} the unsprung member, and the spring constant K ₁ of the unsprung member. Q ₀ = α · M ₁ · q ₁ ″ / K ₁ ... (Formula A)
However, α is a vehicle road surface displacement estimation apparatus according to claim 1, wherein α is a correction coefficient set in advance. The vehicular road surface displacement estimation device according to claim 2, comprising storage means in which the mass M _{1 of} the unsprung member, the spring constant K ₁ of the unsprung member, and the correction coefficient α are stored in advance for each vehicle type.   The vehicle road surface displacement estimation apparatus according to claim 2, further comprising a correction unit that changes a value of the correction coefficient α in accordance with a running state of the vehicle.

Images