JPH068106A - Adaptive control system and state judgment device - Google Patents

Abstract

(57) [Abstract] [Purpose] For the cutting phenomenon of a processing machine, determine whether or not there is an abnormality in the state at high speed and with high accuracy, and perform optimal control. [Structure] The tool 9 and the work piece 1 are constituted by the force sensors 8 and 13.
The processing reaction force of No. 4 is detected (S1), the center of gravity of the vector locus for one cycle of the value is obtained (S2), and the variance of the distance from the center of gravity to each point on the vector locus is obtained (S2).
3). Next, the variance value is compared with the set reference value, and when the variance value exceeds the reference value (S4ye
s), assuming that a machining abnormality has occurred, the machining conditions such as the rotation speed of the tool 9 and the feed speed of the workpiece 14 are changed to avoid the machining abnormality (S5).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an adaptive control system and a state determination device, which is suitable for determining whether or not a processing state is normal from the detection information of a force sensor installed on a tool or a workpiece in a processing machine. It is used when trying to control the processing.

[0002]

2. Description of the Related Art In recent years, in processing machines, so-called processing has become more intelligent, and the state of processing is detected by a sensor such as a six-component force table, and correction is made in real time according to the processing state, thereby improving processing. Higher precision is being attempted. For example, if chattering occurs during cutting with an end mill, the processing accuracy will be impaired. Therefore, in order to detect chatter, conventionally, the magnitude of the cutting force is detected by a force sensor installed on the tool side or the workpiece side, and the detected value is FFT.
(Fast Fourier transform) may be performed to extract and determine the vibration frequency component of chatter. FIG. 13 shows a change in the cutting force (X-axis direction) measured in the actual groove cutting process. The processing conditions in this case are as follows.

Tool: Square end mill (2 blades) Tool diameter: φ5 mm Twist angle: 30 ° Workpiece: S45C Spindle speed: 2200 rpm Feed rate: 60 mm / min Feed direction: Y-axis positive direction Depth of cut: 2.5 mm Machining type: Groove cutting

FIG. 14 shows a frequency distribution (power spectrum) obtained by performing a fast Fourier transform on the change in cutting force shown in FIG. FIG. 15 shows changes in the cutting force in the Y-axis direction during the cutting shown in FIG. FIG. 16 shows a frequency distribution (power spectrum) obtained by performing a fast Fourier transform on the change in cutting force shown in FIG. In these FIGS.
It is in the state where normal cutting is being performed.
In No. 5, the peak value of the cutting force is the rotation cycle of the tool (about 0.02
It appears twice every 7 seconds). Also, FIGS.
In FIG. 6, it can be seen that the components of multiples of the frequency of the peak value appearing in FIGS. 13 and 15 stand out.

FIGS. 17 and 19 show changes in the cutting force when the chattering phenomenon occurs in the cutting under the same conditions as described above in the X-axis and Y-axis directions, respectively. 18 and 20 show frequency distributions (power spectra) obtained by fast Fourier transforming the changes in the cutting force in FIGS. 17 and 19, respectively. 17 and 19, the change component of the cutting force due to the chattering phenomenon is superimposed on the change of the cutting force due to normal cutting. Further, in FIGS. 18 and 20, as is clear from comparison with FIGS. 14 and 16, in addition to the frequency component corresponding to the original peak value of the cutting force, a large amount of the chattering phenomenon appears (especially about 800 Hz). Around). In this way, the occurrence of the chattering phenomenon can be detected by performing the fast Fourier transform on the detected change in the cutting force.

[0006]

By the way, as a method of judging the occurrence of the chattering phenomenon from the waveform of the cutting force and the result of the fast Fourier transform of the cutting force, the simplest method is to visually check the human. However, since there is a demand for automation and unmanned processing, it must be determined by a computer. It is possible to use a pattern recognition method as a specific determination method therefor, but the image processing time becomes long and it is impossible to make determination in real time during processing.

It is also possible to pass the result of the fast Fourier transform through a low-pass filter and take out only the required waveform to make a decision, but similarly the processing time becomes long and it is impossible to make a decision in real time. Moreover, as is well known, the result of the fast Fourier transform has an error, and the processing / judgment is performed while the error is included.
It may be insufficient in terms of accuracy.

As described above, conventionally, a change in cutting force during machining can be detected by using a six-component force sensor or the like, but sufficient processing speed and accuracy can be obtained at the stage of judging the presence or absence of machining abnormality from the detected value. The determination method to be obtained has not been established, and there was a problem in terms of practicality. The present invention has been made to solve the above problems, and an object of the present invention is to determine the presence / absence of a state abnormality in various phenomena including a cutting phenomenon of a processing machine at high speed and with high accuracy, and to optimize the situation. An object of the present invention is to provide an adaptive control system and a state determination device capable of controlling processing.

[0009]

In order to achieve the above object, in the first aspect of the present invention, the detection information of a sensor installed on the fluctuating target side or / and the means side for operating the target is pre-input. It is characterized in that the condition of the target is judged based on the predicted value obtained from the operating condition, and the operating means is controlled so that the state of the target and / or the state of the operating means becomes more appropriate from the result of the judgment. . A second invention is characterized in that, in the first invention, a predicted value is obtained from an operation condition input in advance by using a physical model concerning an object.

According to a third aspect of the present invention, in the first or second aspect of the present invention, the operation status determination result is accumulated as a rule regarding the operation, knowledge regarding the operation means and the operation target, and knowledge regarding the operation condition and the operation result. The feature is that the user learns by feeding back to the database and the operation condition determination unit. A fourth aspect of the present invention determines a machining situation based on detection information of a sensor installed on a tool side or / and a workpiece side of a machining machine and a predicted value obtained from a machining condition input in advance, and makes the determination. From the result, the processing machine is controlled so that the state of the processing phenomenon and / or the state of the processing machine becomes more appropriate.

A fifth aspect of the invention is characterized in that, in the fourth aspect of the invention, a predicted value is obtained from a previously input machining condition using a physical model of machining. 6th invention, 4th
Alternatively, in the fifth invention, the determination result of the processing situation is
It is characterized in that it learns on its own by feeding back to the database for accumulating knowledge about machining rules, knowledge about tools and workpieces, knowledge about machining conditions and machining results, and the machining condition determination unit.

According to a seventh aspect of the present invention, a sensor for detecting a state quantity of an object that changes periodically and time as a parameter are used to draw the detected state quantity as a vector locus in a one-dimensional or multidimensional space. A means for obtaining the center of gravity of a figure surrounded by the trajectory, a means for obtaining the variance of the distance from the center of gravity to each point on the trajectory of the vector, and comparing the variance value with a reference value to determine whether there is an abnormality in the target. And means.

According to an eighth invention, in the seventh invention, a database accumulating knowledge about a physical law governing an object and a record of the behavior of the object detected in the past,
And a means for calculating a reference value for determination by constructing a physical model of the target while reproducing or predicting the state change of the target at the present time point and after that while referring to the database.

According to a ninth aspect of the present invention, a sensor for detecting a machining reaction force generated between a tool of a machining machine and a workpiece, which changes cyclically, and time are used as parameters. The detected machining reaction force value is 1
Draw as a vector trajectory on a dimensional or multidimensional space,
Means for obtaining the center of gravity of the figure surrounded by the trajectory, means for obtaining the variance of the distance from the center of gravity to each point on the trajectory of the vector, and comparing the variance value with a reference value to determine the presence or absence of machining abnormality. And means.

A tenth aspect of the present invention relates to the ninth aspect of the present invention, which is related to knowledge about a rule governing a machining phenomenon between a tool and a workpiece in a machining machine and a change in a machining reaction force detected for each machining condition in the past. A means for calculating a reference value for judgment by constructing a database that stores records and a physical model related to machining between a tool and a workpiece while referring to the database and reproducing or predicting machining phenomena at the present time and thereafter. It is characterized by having and.

[0016]

According to the first aspect of the present invention, the situation of the target is based on the detection information of the sensor installed on the fluctuating target side or / and the means side for operating the target and the predicted value obtained from the operation condition input in advance. Is determined, and the operation means controls the state of the object and / or the state of the operation means to be more appropriate from the result of the determination. In the second invention, the predicted value is obtained from the operation condition input in advance by using the physical model regarding the object.

In the third aspect of the present invention, the operation status judgment results are stored in a database that stores rules regarding operations, knowledge about operation means and operation targets, operation conditions and knowledge about operation results, and an operation condition determining unit. Learn by yourself by being fed back. In the fourth aspect of the invention, the machining situation is judged based on the detection information of the sensor installed on the tool side or / and the workpiece side of the machining machine and the predicted value obtained from the machining condition input in advance. From the judgment result, the processing machine is controlled so that the state of the processing phenomenon and / or the state of the processing machine becomes more appropriate.

In the fifth aspect of the invention, the predicted value is obtained from the previously input processing conditions using the physical model of processing. In the sixth aspect of the present invention, the result of the determination of the machining situation is fed back to the rule regarding machining, the knowledge regarding the tool and the workpiece, the database accumulating the knowledge regarding the machining condition and the machining result, and the determining unit for the machining condition. Learn by yourself.

In the seventh invention, when the state quantity of the object which changes periodically is detected by the sensor, one kind or two or more different kinds of state quantities are one-dimensional or multi-dimensional with time as a parameter. It is drawn as a trajectory of a vector in a dimensional space, and the center of gravity of a figure surrounded by the trajectory is obtained.
Then, the variance of the distance from the center of gravity to each point on the trajectory of the vector is obtained, and the value of the variance is compared with the reference value to determine whether the target state is abnormal.

In the eighth aspect of the invention, the physical model of the object is constructed with reference to the knowledge of the physical laws governing the object accumulated in the database and the record of the behavior of the object detected in the past. The reference value for determination is calculated by reproducing and predicting the subsequent change in the state of the target.

According to the ninth aspect of the invention, when the processing reaction force generated between the tool of the processing machine and the workpiece which changes cyclically is detected by the sensor, one or two or more different processing reaction forces are detected. The force value is drawn as a vector locus in one-dimensional or multidimensional space with time as a parameter, and the center of gravity of the figure surrounded by the locus is obtained. Next, the variance of the distance from the center of gravity to each point on the vector locus is obtained, and the variance value and the reference value are compared to determine whether or not there is an abnormality in the machining.

In the tenth aspect of the invention, knowledge of the law governing the machining phenomenon between the tool and the workpiece in the machining machine accumulated in the database and the variation of the machining reaction force detected for each machining condition in the past. A physical model relating to machining is constructed with reference to the record relating to machining, and a reference value for determination is calculated by reproducing and predicting machining phenomena at the present time point and thereafter.

[0023]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an overall configuration when the present invention is applied to machining center machining. In the figure, the processing condition determination unit 1 determines the processing conditions (material of the work, type of tool, cutting speed, feed speed, depth of cut) with reference to the database 22 based on the input design information. Then, the specific operation amount of each actuator for operating the machining center 5 is determined and sent to the machining machine controller 2, the machining physical model 3, and the sensor information predictor 4.

The machining machine controller 2 drives the rotating mechanism 7 and the moving mechanism 6 of the tool 9 and the moving mechanism 11 of the workpiece 14 based on the manipulated variable to control the machining of the workpiece 14 by the tool 9. The drive amount of the rotation mechanism 7 and the movement mechanisms 6 and 11 is corrected by the operation signal from the adaptive controller 21. The force sensor 8 installed in the tool 9 detects a processing reaction force generated in the tool 9 during processing and sends a detection signal proportional to the processing reaction force to the strain amplifier 16. The force sensor 8 is composed of a six-component force sensor or the like and detects the processing reaction force generated in the tool 9 as a six-axis component.

Similarly, the force sensor 13 installed on the side of the workpiece 14 detects a processing reaction force generated on the workpiece 14 during processing, and outputs a detection signal proportional to the processing reaction force to the strain amplifier 1.
Send to 7. A fail-safe mechanism 15 is provided between the rotation mechanism 7 and the force sensor 8, and a moving mechanism 11 and the force sensor 1 are provided.
A fail-safe mechanism 12 is installed between each of the 3 and 3, and when an excessive load is applied to the tool 9 and the workpiece 14 during machining, the tool 9, the workpiece 14 and the force sensor 8,
Prevent damage to 13. The distortion amplifiers 16 and 17 respectively amplify the detection signals and send them to the sensor information acquisition device 18.

The sensor information acquisition device 18 sends each detection signal to A
/ D-converted and sent to the state determiner 19 at a predetermined timing. On the other hand, a physical model 3 of machining corresponding to the machining center 5 is created from the operation amount output from the machining condition determining unit 1, and the machining phenomenon between the tool 9 and the workpiece 14 is reproduced or predicted. . The model data obtained by the processing physical model 3 is sent to the sensor information predictor 4.
The sensor information predictor 4 predicts the detection values of the force sensors 8 and 13 from the operation amount and the model data, and sends the prediction value to the state determiner 19.

Here, the state determiner 19 compares the input detected values of the force sensors 8 and 13 with the predicted value, and if the detected value exceeds the allowable range of the predicted value, it is determined that the machining is abnormal. Then, the state determination signal is sent to the adaptive controller 21 and the database 22. Even if the allowable range is not exceeded, how appropriate the current machining state is is determined by the magnitude of the value. In the state determiner 19,
Although the detection values of both the input force sensors 8 and 13 are used as the determination reference, it is also possible to use only the detection values of one sensor to make the determination.

When the state determination signal is input, the adaptive controller 21 creates an operation signal for improving the processing state and sends it to the processing machine controller 2. The database 22 stores in advance rules regarding machining, knowledge regarding tools and workpieces, knowledge regarding machining conditions and machining results, and further acquires and accumulates knowledge regarding machining from the state determination signal from the state determiner 19. Although not shown, the knowledge in the database 22 is also used by the adaptive controller 21 and the physical model 3 for machining, which is not shown, in addition to being called by the machining condition determining unit 1.

The rules regarding machining in the database 22 are, for example, when the chattering phenomenon occurs during machining, the rule is to take concrete measures when an abnormality occurs, such as reducing the feed rate and increasing the rotation speed. Is. Further, the knowledge about the tool and the workpiece is a usage condition for performing appropriate machining such as the type of machining that can be performed by the tool and the machining range by the combination of the tool and the workpiece.

Similarly, the knowledge about the machining conditions and the machining results is the knowledge already known empirically such as "a chatter phenomenon occurs with a certain tool at any rotation speed and feed rate". In addition to that, the knowledge actually gained by experience at the machining center 5 is also accumulated. That is, when the result of processing under a certain processing condition is defective, the processing result is fed back and the processing condition is corrected. In this way, the learning is repeated as the processing experience is gained, and the accuracy of the accumulated knowledge is improved. In addition, rules regarding processing are also accumulated.

Next, a method for determining a machining abnormality will be specifically described. First, the value of the machining reaction force generated between the tool 9 and the workpiece 14 when the workpiece 14 is machined is obtained by the physical model 3 based on the machining condition determined by the machining condition determination unit 1. FIG. 2 shows changes in the machining reaction force obtained based on the physical model 3 of machining under the following machining conditions as two-direction components on the X and Y planes perpendicular to the axial direction of the tool 9. , Is a vector locus with time as a parameter.

Tool: Square end mill (2 blades) Tool diameter: φ5 mm Twist angle: 30 ° Workpiece: S45C Spindle speed: 2200 rpm Feed rate: 60 mm / min Feed direction: Y-axis positive direction Depth of cut: 2.5 mm Machining type: Groove cutting

As is apparent from the figure, the change in the machining reaction force has a periodicity and the locus of its vector is a closed curve. Since the tool 9 is a double-edged blade, the closed curve can be traced on a closed curve in one half rotation. Make one lap. Next, the center of gravity of the figure enclosed by the closed curve drawn by the vector locus and the variance of the vector locus are obtained. FIG. 3 is an enlarged view of the vector trajectory of FIG. 2. After obtaining the center of gravity (Gx, Gy), each point (Fx1, Fy1) ...
The variance of the distance to (Fxi, Fyi) ...
The value of this variance is a small value when the vector locus is close to a circle and is a smooth shape, and is large on the contrary when the shape has many undulations different from the circle, and the change in the outer shape of the rotating figure is estimated. Can be used as a guide.

Next, the values of the processing reaction forces detected from the actual processing will be compared. FIG. 4 shows the tool 9 when the workpiece 14 is actually machined under the same machining conditions as the machining shown in FIG.
The detected value of the processing reaction force generated between the workpiece and the workpiece 14 is represented as a vector locus. Tool 9 in the figure
1 is represented by one rotation, and the vector trajectory is a double elliptical trajectory. Compared with the locus of the model shown in FIG. 2, a slight undulation is seen on the outer shape, which indicates a normal working state. The center of gravity and the variance are similarly obtained from this trajectory.

FIG. 5 shows the locus of the vector when the chattering phenomenon occurs in the processing shown in FIG. As is clear from the figure, when the chattering phenomenon occurs, the amplitude appears in the locus and the value of dispersion increases. Also for this locus, the center of gravity and amplitude for one rotation are obtained. FIGS. 6 and 7 show the temporal change of the dispersion obtained from the values actually measured in this way. FIG. 6 shows the change in dispersion in the normal processing shown in FIG. 4 in comparison with the dispersion value (161N ² ) obtained from the physical model.

FIG. 7 shows the change in dispersion due to the abnormal processing shown in FIG. As is clear from a comparison between the two figures, in the normal case, the variance value is within several times the value obtained from the physical model, and the degree of change is moderate.
On the other hand, in an abnormal case, the chattering phenomenon itself appears as a waveform, and the dispersion value is extremely larger than the model value, and the change is drastic. As a result, it is possible to determine whether the machining is normal or abnormal based on the magnitude of the variance and the degree of change. That is, the ratio of the variance between the measured value and the model value is set in advance, the variation of the variance value obtained during machining is monitored, and if it exceeds the set ratio, it is determined that machining is abnormal. The values of the center of gravity and the variance are for one half rotation of the tool 9,
Alternatively, it can be calculated for each detection data for two rotations.

FIGS. 8 and 9 show changes in the variance value in the normal processing and the abnormal processing in other processing examples, respectively. In this processing example, the feed rate is 120 mm / min, and the other processing conditions are the same as those in the processing example described above. By setting the feed rate to 120 mm / min, the dispersion value obtained from the model is 349 N ² . 10 and 11 respectively show a normal case and an abnormal case in other processing examples. In this processing example, the feed rate is 90 mm / min
Other processing conditions are the same as the processing example described above. By setting the feed rate to 90 mm / min, the dispersion value obtained from the model is 251 N ² .

In each of the machining examples described above, the center of gravity and the variance were obtained from the two-dimensional vector locus, focusing on the biaxial component where the change in the machining reaction force is remarkable. If it is conspicuous only for one axis, it is possible to determine the machining abnormality by the same process even for only one axis. Further, if the change cycle is the same, it is possible to determine the machining abnormality by obtaining the center of gravity and the variance as the trajectory of the vector in the multidimensional space using the detection values of three or more axes.

FIG. 12 is a flow chart systematically showing the determination processing of the above-mentioned embodiment, and the following description will be made with reference to the drawing. First, the force sensors 8 and 13 detect the processing reaction force generated on the tool 9 and the workpiece 14 during processing (S
1). Next, the center of gravity of the figure drawn by the trajectory when the one cycle of the detected processing reaction force value is represented as the trajectory of the vector is determined (S2), and the variance of the distance from the center of gravity to each point on the trajectory of the vector is determined. (S3).

The obtained dispersion value is compared with the reference value which is set in advance according to the processing conditions. When the obtained dispersion exceeds the reference value (S4 yes), a processing abnormality has occurred. As a matter of course, the machining conditions such as the rotational speed of the tool 9 and the feed speed of the workpiece 14 are changed to avoid abnormal machining (S
5). In this way, even if a chattering phenomenon occurs during machining, the tool 9 and the workpiece 14 are prevented from being damaged by being immediately detected and returned to the normal machining state, and the life of the tool 9 is increased. However, the yield of processed products is also improved. In the embodiment, the presence / absence of abnormality is determined by paying attention to the change in the two-direction components of the generated processing reaction force of the X-axis and the Y-axis. Even if there is, it is similarly possible, and the measured values are combined so as to make the determination most easily according to the characteristics of the actual determination target.

Further, in the embodiment, the chatter phenomenon is explained as the machining abnormality, but other machining abnormality, for example, the machining abnormality due to wear or damage of the tool 9 can be similarly detected. Furthermore, as another processing machine,
It can also be applied to lathes, grinders, drilling machines, etc. Further, it is also applicable to objects other than processing machines. That is,
If the measured value is the state quantity of the object that changes with time and has periodicity, the measured value is used as the vector locus to determine the center of gravity and variance, and it can be compared with the reference value to determine whether there is an abnormality. Is. As the reference value, a physical model may be used, or a fixed value may be set in advance.

[0042]

As described above, according to the first invention,
It is possible to more appropriately control the state of the target or the state of the operating means by determining and controlling the fluctuating target or the state of the means for operating the target based on the predicted value from the previously input operating conditions. it can. According to the second aspect of the present invention, the predicted value is calculated from the preliminarily input operation condition using the physical model regarding the object, whereby the predicted value can be calculated at high speed and with high accuracy, and more appropriate control can be performed.

According to the third aspect of the present invention, the judgment result of the operation status is fed back to the database or the like and learned by itself, whereby more appropriate control can be performed as the operation experience is gained. According to the fourth aspect of the present invention, the state of the tool or the workpiece is more appropriately determined by determining and controlling the situation of the tool or the workpiece during the machining based on the predicted value from the machining conditions input in advance. Can be controlled.

According to the fifth aspect of the invention, the predicted value is calculated from the previously inputted processing conditions by using the physical model of processing, whereby the calculation of the predicted value becomes fast and highly accurate, and the processing accuracy is improved. Can be made. According to the sixth aspect, the determination result of the machining status is fed back to the database or the like and learned by itself, so that more accurate machining can be performed as the machining experience is gained.

According to the seventh aspect of the present invention, the center of gravity and the variance of the state quantity of the object detected by the sensor are obtained, and the obtained value of the variance is compared with the reference value to judge whether or not the state of the object is abnormal. Thus, the determination can be made only if the detected data is for one cycle in which the center of gravity and the variance are obtained. Therefore, it is possible to determine the presence or absence of abnormality by a short-time calculation, and it is possible to determine the target state in real time. According to the eighth aspect, since the reference value is calculated based on the physical model constructed while referring to the database, the reference value is always the optimum value according to the object, and the determination result is highly accurate.

According to the ninth aspect of the invention, the center of gravity and the dispersion of the value of the processing reaction force generated between the tool of the processing machine and the workpiece detected by the sensor are obtained, and the obtained dispersion value and the reference value are used. By comparing whether there is an abnormality in processing,
The determination can be made if the detected data amount is only one cycle in which the center of gravity and the variance are obtained. Therefore, it becomes possible to determine the presence or absence of abnormality by a short-time calculation, and it becomes possible to determine the processing state of the processing machine in real time. According to the tenth invention, since the reference value is calculated based on the physical model of processing constructed while referring to the database, the reference value is always the optimum value according to the processing conditions, and the determination result is highly accurate. Become.

[Brief description of drawings]

FIG. 1 is a diagram showing an overall configuration of an embodiment of the present invention.

FIG. 2 is a graph showing a machining reaction force obtained based on a physical model of machining as a vector locus.

FIG. 3 is a graph showing the trajectory of the vector of FIG. 2 in an enlarged manner and showing the center of gravity of the trajectory and the distance from the center of gravity to the trajectory of the vector.

FIG. 4 is a graph showing a machining reaction force actually generated as a vector locus.

FIG. 5 is a graph showing a locus of a vector when a machining abnormality occurs.

FIG. 6 is a graph showing changes in dispersion during normal processing.

FIG. 7 is a graph showing changes in dispersion due to abnormal processing.

FIG. 8 is a graph showing another example of changes in dispersion during normal processing.

FIG. 9 is a graph showing another example of changes in dispersion due to abnormal processing.

FIG. 10 is a graph showing another example of changes in dispersion during normal processing.

FIG. 11 is a graph showing another example of changes in dispersion due to abnormal processing.

FIG. 12 is a flowchart systematically showing the determination processing of the embodiment.

FIG. 13 is a graph showing changes in the X-axis component of the actually measured cutting force.

FIG. 14 is a graph showing the result of fast Fourier transform of the change in cutting force in FIG.

FIG. 15 is a graph showing changes in the Y-axis component of the actually measured cutting force.

16 is a graph showing the result of fast Fourier transform of the change in cutting force in FIG.

FIG. 17 is a graph showing changes in the X-axis component of the cutting force when a cutting abnormality occurs.

18 is a graph showing the result of fast Fourier transform of the change in cutting force in FIG.

FIG. 19 is a graph showing changes in the Y-axis component of the cutting force when a cutting abnormality occurs.

20 is a graph showing the result of fast Fourier transform of the change in cutting force in FIG.

[Explanation of symbols]

 1 Machining condition determination unit 2 Machining machine controller 3 Physical model of machining 4 Sensor information predictor 5 Machining center 6 Moving mechanism 7 Rotating mechanism 8 Force sensor 9 Tool 11 Moving mechanism 12 Fail-safe mechanism 13 Force sensor 14 Workpiece 15 Fail Safe mechanism 16 Distortion amplifier 17 Distortion amplifier 18 Sensor information acquisition device 19 State determination device 21 Adaptive controller 22 Database

Claims (10)

[Claims] 1. A situation of a target is judged based on detection information of a sensor installed on a fluctuating target side or / and a means side for operating the target and a predicted value obtained from an operation condition inputted in advance, and The state of the target or /
And an adaptive control system which controls the operating means so that the state of the operating means becomes more appropriate. 2. The adaptive control system according to claim 1, wherein a predicted value is obtained from a previously input operation condition by using a physical model regarding the object. 3. The adaptive control system according to claim 1, wherein the operation status judgment result is a database in which rules regarding operations, knowledge about operation means and operation targets, knowledge about operation conditions and operation results are accumulated, and An adaptive control system characterized in that feedback is provided to the operation condition determination unit to learn by itself. 4. The machining situation is judged based on the detection information of the sensor installed on the tool side or / and the workpiece side of the processing machine and the predicted value obtained from the machining conditions inputted in advance, and the judgment result thereof. An adaptive control system for controlling a processing machine so that the state of a processing phenomenon and / or the state of a processing machine becomes more appropriate. 5. The adaptive control system according to claim 4, wherein a predicted value is obtained from a machining condition input in advance using a physical model of machining. 6. The adaptive control system according to claim 4 or 5, wherein the determination result of the machining status is a database that stores rules regarding machining, knowledge about tools and workpieces, knowledge about machining conditions and machining results, and An adaptive control system characterized by feedback to a processing condition determination unit for self-learning. 7. A sensor for detecting a state quantity of an object that changes periodically, a time as a parameter, and a detected state quantity drawn as a vector locus in a one-dimensional or multidimensional space and surrounded by the locus. The means for obtaining the center of gravity of the figure, the means for obtaining the variance of the distance from the center of gravity to each point on the vector trajectory, and the means for comparing the variance value with the reference value to determine whether there is any abnormality in the target. A state determination device characterized by being provided. 8. A database in which knowledge about physical laws governing an object and records of behavior of the object detected in the past are accumulated, and a physical model of the object is constructed by referring to the database to present and subsequent object states. The state determination device according to claim 7, further comprising: a unit that calculates a reference value for determination by reproducing or predicting a change. 9. A sensor for detecting a machining reaction force generated between a tool of a machining machine and a workpiece which changes periodically, and a detected machining reaction force value as one-dimensional or multi-dimensional with time as a parameter. A method of drawing the trajectory of a vector in space and finding the center of gravity of the figure surrounded by the trajectory, a method of finding the variance of the distance from the centroid to each point on the vector trajectory, and comparing the variance value with a reference value Then, a state determination device comprising: means for determining the presence or absence of processing abnormality. 10. A database accumulating a record of knowledge about a law governing a machining phenomenon between a tool and a workpiece in a machining machine and a record of a variation of a machining reaction force detected for each machining condition in the past, and the database. However, a means for calculating a reference value for judgment by constructing a physical model for machining between the tool and the workpiece and reproducing or predicting the machining phenomenon at the present time and thereafter is provided. The state determination device according to claim 9.