JPH06242833A - Automatic adjusting servo controller - Google Patents

Abstract

PURPOSE:To provide an automatic adjusting servo controller capable of always performing control by the optimum control parameter conforming to the change of a load. CONSTITUTION:This controller is provided with a load change estimation part 9, a control parameter storage part 10 which stores the load of every kind of size and the optimum control parameter in accordance with the load, and a control mechanism 12 which controls the whole system, and uses the optimum parameter when a reference load is applied when a servo operation is started, and regulates the actual size of the load by integrating difference between a deviation counter curve when the servo operation is started and that when the optimum parameter is used with a standard load for a fixed time by the load change estimation part 9, and uses the optimum control parameter by reading out from the control parameter storage part 10 according to the actual size of a regulated load when the fixed time elapses.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a servo control device used in a positioning system for a robot or the like, and more particularly to an automatic adjustment servo control device for automatically adjusting control parameters and the like in response to changes in the load applied to a motor. It is about.

[0002]

2. Description of the Related Art The configuration and operation of a conventional servo control device used in a positioning system for a robot or the like will be described with reference to FIG.

In FIG. 19, in order to control the motor 6 so that the controlled object coincides with the target position at each time,
The command control unit 1 instructed the target position outputs the position command at each time based on the separately set acceleration and maximum speed. The deviation counter 2 receives the position command and the position data from the position detector 7, calculates these deviations, and outputs the calculated deviations. The position amplifier 3 receives the output of the deviation counter 2 and receives the position amplifier proportional gain Kpp.
Is calculated proportionally and the speed command value is output. Speed amplifier 4
, The proportional / integral calculation is performed according to the difference between the speed command value received from the position amplifier 3 and the speed data from the speed detector 8 to obtain the speed amplifier proportional gain Kvp and the speed amplifier integral gain Kvi / S. Is output. S represents the Laplace transform. The drive circuit 5 uses the speed amplifier proportional gain Kvp
The motor 6 is driven based on and the speed amplifier integral gain Kvi.

Since the inertial load, the frictional resistance, etc. differ depending on the positioning system to be controlled, in order to obtain the target control characteristics, the position amplifier proportional gain Kpp of the position amplifier 3 and
It is necessary to adjust the calculation of the speed amplifier proportional gain Kvp of the speed amplifier 4 and the speed amplifier integral gain Kvi.
In the conventional example, an expert performs this adjustment. The main method of this adjustment is to move the system under the actual operating conditions, adjust the calculation of each of the above control parameters by turning the volume while observing the operating curve with an oscilloscope, and the speed control loop, There is a method in which stepwise command values are given in the order of the position control loop and the calculation of each of the above-mentioned control parameters is adjusted by the response characteristic thereof.

[0005]

However, the configuration of the above conventional example has a problem that it takes a lot of time to adjust the calculation of the control parameters. Therefore, after adjusting the above control parameters, it is difficult to adjust the control parameters during operation, and in a system in which the load fluctuates significantly during operation, the target control characteristics can be obtained in all states. However, there is a problem in that, in order to achieve this, it takes an extremely huge amount of time for adjustment each time.

An object of the present invention is to solve the above-mentioned problems and to provide an automatic adjustment servo control device which always controls with an optimum control parameter in response to a change in load.

[0007]

In order to solve the above-mentioned problems, an automatic adjustment servo control apparatus of the present invention controls a motor to match the movement of an object to be servo-controlled with a target position at each time. A command control unit that outputs a position command value, a deviation counter that detects a difference between the position command value and the actual position and outputs a deviation counter curve, and a position control loop and a speed control loop that operate based on control parameters are provided. In a servo control device having: a load change estimation unit that defines a load size from the deviation counter curve; and a control parameter storage unit that stores preset load sizes and optimum control parameters corresponding thereto. , Control the entire control unit to use the preset optimum parameters for standard load when starting the servo operation. The load change estimation unit integrates the difference between the deviation counter curve at the start of servo operation and the reference model of the deviation counter curve when optimum parameters are used with a preset standard load for a certain period of time, and the result is calculated. Optimal control corresponding to the magnitude of the actual load from the control parameter storage unit based on the magnitude of the specified actual load at the end of the fixed time, by defining the magnitude of the actual load. And a control mechanism for reading out parameters and controlling using the parameters.

Further, the automatic adjustment servo control device of the present invention is
In order to solve the above-mentioned problem, the overshoot amount and the settling time are calculated from the deviation counter curve, and the control parameter auto-tuning unit that adjusts the control parameter by fuzzy reasoning from the calculation result, and the entire control device is controlled. The control parameter auto-tuning unit is caused to calculate an optimum control parameter at a standard load, the optimum control parameter is stored in a control parameter storage unit, and the servo operation is performed at the standard load using the optimum control parameter. The deviation counter curve for a fixed time from the start of servo operation is obtained as a reference model of the deviation counter curve, this reference model is stored in the control parameter storage unit, and only the load is sequentially changed to various loads. In this case, load changes for a certain time from the start of servo operation. The estimation unit is made to obtain an integral value of the difference between the deviation counter curve at the start of the servo operation and the reference model of the deviation counter curve, and at the same time, the control parameter auto-tuning unit is made to calculate the optimum control parameter at the various loads. It is preferable to have a control mechanism for storing a combination of the integrated value of the difference and the optimum control parameter in the control parameter storage unit at the end of the fixed time.

[0009]

In the conventional example, there is no means for making a one-to-one correspondence between the changing load and the optimum control parameter corresponding thereto, and the control parameter can be adjusted to the optimum value in response to the change in the load. However, in the automatic adjustment servo control device of the present invention, a means for associating the changing load with the corresponding optimum control parameter on a one-to-one basis is obtained to solve the above problem. In response to changes in load during operation, it is possible to always control with optimum control parameters, and it has the following effects.

In the automatic adjustment servo controller of the present invention, when the load changes during operation, the automatic tuning is performed in order to specify the magnitude of the load corresponding to each optimum control parameter for each load on a one-to-one basis. At the start of the servo operation, the control mechanism that performs the operation uses the preset optimum parameters at the standard load, and the load change estimation unit displays the deviation counter curve at the start of the servo operation for a certain period of time from the start of the servo operation. , The difference between the deviation counter curve and the reference model when the optimum parameter is used with a preset standard load is integrated, and the result determines the actual load magnitude at the start of servo operation. The magnitude of the load thus defined has a one-to-one correspondence with the optimum parameter for the magnitude of the specified load.

Therefore, if the control parameter storage section stores the magnitudes of various loads set in advance according to the above regulation and the optimum control parameters corresponding thereto, the magnitudes of the above-mentioned prescribed loads can be stored. The control mechanism reads the optimum parameters from the control parameter storage unit, whereby the optimum control parameters corresponding to the changing load can be obtained.

Further, in the automatic adjustment servo control device of the present invention, the magnitudes of various loads set by the above-mentioned prescribed method and the optimum control parameters corresponding thereto are automatically calculated and stored in the control parameter storage section. In order to do so, an overshoot amount and a settling time are calculated from the deviation counter curve, and a control parameter auto-tuning unit that adjusts the control parameter by fuzzy reasoning from the calculation result is added, and the control mechanism for performing the auto-tuning is The parameter auto-tuning unit calculates the optimum control parameter at the standard load, the optimum control parameter at the standard load is stored in the control parameter storage unit, and the servo operation is performed at the standard load using the optimum control parameter. , Deviation of load change estimation unit for a fixed time from the start of servo operation The uncertain curve is calculated as a reference model of the deviation counter curve, this reference model is stored in the control parameter storage unit, and only the load is sequentially changed to various loads. For these various loads, the constant is maintained from the start of servo operation. For a certain time, the load change estimation unit is made to obtain an integral value of the difference between the deviation counter curve at the start of servo operation and the reference model of the deviation counter curve, and at the same time, the control parameter auto-tuning unit is optimized for the various loads. A control parameter is calculated, and at the end of the certain period of time, a combination of the integrated value of the difference and the optimum control parameter is stored in the control parameter storage unit.

According to the above, according to the present invention, even if the load changes, the control parameters used until then are optimized by the auto tuning until the maximum speed is reached by starting the servo control operation. By changing to the control parameter, the servo control is always performed with the optimum control parameter.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The construction of an embodiment of the automatic adjustment servo control system of the present invention will be described with reference to FIGS.

In the orthogonal robot shown in FIG. 2, the Y-axis table 13 has a Y-axis motor 14 and a Y-axis ball screw 15, and the X-axis table 16 is moved along the Y-axis table 13 in the Y-axis direction. , Position it at any position.
The X-axis table 16 has an X-axis motor 17 and an X-axis ball screw 18. The movable body 19 is moved along the X-axis table 16 in the X-axis direction and positioned at an arbitrary position. A work head 20 is attached to the movable body 19. The working head portion 20 has a predetermined distance D in the X-axis direction.
Position recognition camera 2 for recognizing work positions
1 and a work tool 22 for performing a predetermined work are fixed. In the example of FIG. 2, the axis of the position recognition camera 21 and
The axis of the working tool 22 and the axis of the working tool 22 are accurately arranged on a line along the X axis direction. The work head unit 20 is installed so as to be movable in the X-axis direction with high accuracy with respect to the movable body 19, and its moving range substantially corresponds to the distance D between the position recognition camera 21 and the work tool 22. There is. Further, there is provided a precision positioning means 25 including a precision motor 23 and a precision ball screw 24 for moving the work head portion 20 to perform highly accurate positioning.

FIG. 1 is a block diagram of this embodiment used for servo control of the Y-axis table 13 shown in FIG. still,
X-axis table 16 and working head unit 20 (hereinafter referred to as H-axis).

The servo control device of 1) is also used in the same manner. However, since these operations are the same as those of the Y-axis table 13, the description thereof will be omitted.

In FIG. 1 and FIG. 2, the command controller 1 sets the acceleration α separately set so that the motor 6 is controlled so that the Y-axis table 13 which is the servo control object coincides with the target position at each time. And the maximum speed, the position command at that time is output for each sampling cycle. The deviation counter 2 receives the above-mentioned position command and the position data of the Y-axis table 13 from the pulse generator of the position detector 7, calculates these deviations, and creates a deviation counter curve as shown in FIG. Then, the calculated deviation counter curve is output. The position amplifier 3 receives the output of the deviation counter curve of the deviation counter 2, multiplies the output by the position amplifier proportional gain Kpp, and outputs it as a speed command signal. The speed amplifier 4 transfers the transfer function Kv according to the difference between the speed command signal received from the position amplifier 3 and the speed data from the speed detector 8.
It calculates based on p + Kvi / S and outputs the result as a torque command signal. Here, S represents the Laplace transform. The drive circuit 5 drives the motor 6 based on the torque command signal.

In this case, depending on the work content, during operation,
In a robot in which the work tool 22 and the like are exchanged, the load applied to the X-axis table 16 and the Y-axis table 13 changes, and therefore, in order to obtain the control characteristic that achieves the target in any state, the load changes. Then, it is necessary to change each control parameter used by the position amplifier 3 and the speed amplifier 4.

For this purpose, first, an overshoot amount and a settling time are calculated from the deviation counter curve, and a control parameter auto-tuning unit 11 for adjusting the control parameter by fuzzy reasoning from the calculation result is provided. A control mechanism 12 in 11 controls the entire apparatus and performs an actual servo operation of moving the X-axis table 16 with a standard load with respect to the Y-axis table 13 of the controlled object, and the control parameter auto-tuning unit 11 In addition, the overshoot amount and the settling time are calculated from the deviation counter curve at that time, and the optimum control parameters are obtained by fuzzy inference from the calculation results, and the optimum control parameters are stored in the control parameter storage unit 10. At the same time, The load change estimation unit 9 is provided with the actual servo operation. A deviation counter curve for a fixed time from the start of the servo operation is created, and this deviation counter curve is used as a reference model of the deviation counter curve when the servo operation is performed at the standard load using the optimum control parameters. To memorize.

Next, for various loads that are expected to actually appear, the servo operation using the optimum control parameters corresponding to the standard load is sequentially performed, and the optimum control parameters in each case are set in the control parameter auto-tuning unit 11. Calculate. At the same time, the load change estimation unit 9
In each of the above cases, a deviation counter curve is created for a fixed time from the start of the servo operation during the actual servo operation, and the integrated value of the difference between this deviation counter curve and the reference model of the deviation counter curve is calculated. And calculate the integrated value
The optimum control parameters in each case are combined and stored in the control parameter storage unit 10. By doing so, the integrated value and the optimum control parameter in each case have a one-to-one correspondence.

Next, the auto tuning operation of the control parameter auto tuning unit 11 and the control mechanism 12 will be described with reference to the flowchart of FIG.

At the start of auto-tuning in step # 1, first, the magnitude of the load is set to an intermediate magnitude of the expected load change (this is the standard load). Then, after giving the target position, acceleration, and maximum speed at each time to the command control unit 1, the acceleration α, the position amplifier proportional gain Kpp, the speed amplifier proportional gain Kvp, and the speed amplifier integral gain Kvi / S (S is Laplace transform). The basic value of () is basically investigated in advance, Kvp is the optimum value examined, and α, Kpp, and Kvi are about 1/3 of the optimum value. To do. However, if the characteristics of the machine are known to some extent in advance, it is possible to set the initial value with the known value. When this is finished, proceed to step # 2.

Standard load tuning is performed in the optimum control parameter calculation of the standard load in step # 2. That is, in the standard load, the servo operation is performed once based on the target position, acceleration, and maximum speed set by the command control unit 1. From the operation result, the control parameter auto-tuning unit 11 performs fuzzy inference, and new control parameters α, Kpp, and K are set so that the control characteristics approach the target.
Change to vp and Kvi. This servo operation and control parameter change operation are replaced with new control parameters α, Kpp,
By repeating the change amounts of Kvp and Kvi within a certain fixed range, the optimum control parameter under that condition is obtained. The obtained optimum control parameter is stored in the control parameter storage unit 10. Then step # 3
Proceed to.

In step # 3, a reference model is obtained. That is, using the optimum control parameters obtained as described above, the servo operation is performed, and for each sampling cycle of position control, a certain fixed time (the time from the start of the servo operation to the time when the command reaches the maximum speed) The deviation counter curve up to 1/3) is stored in the load change estimating unit 9 as a reference model. The value of this reference model is Eref (kT). Where T is the sampling time, k is 0, 1, 2,
3, ... Then, it proceeds to step # 4.

In step # 4, the integral value of the difference between the deviation counter curve under various loads and the reference model is obtained. That is, the magnitude of the load is sequentially changed appropriately, the servo operation is performed using the optimum control parameter for the standard load as the control parameter, and the control parameter auto-tuning unit 11
From the start of the servo operation, the output signal {E (kT)} of the deviation counter 2 is observed at each sampling cycle of position control, the difference from the reference model value Eref (kT) is obtained, and the certain fixed time is reached. The integral value of the difference up to (1/3 of the time until the command reaches the maximum speed) is calculated. If this integrated value is Ei (1), then Ei (1) = Σ {E (kT) −Eref (kT)}. In this case, if the sampling time is 1 ms and the time to reach the maximum speed is 100 ms, 100 m
The above integrated value is calculated until 33 ms, which is 1/3 of s, is reached. Then, it proceeds to step # 5.

In step # 5, optimum control parameters for each load are obtained. Simultaneously with step # 4, automatic adjustment is performed as in the case of standard load, and the optimum control parameters (Kpp, Kvp, Kvi) for each load obtained as a result are obtained.
Ask for. These operations are performed by changing the load in several stages. Then, it proceeds to step # 6.

In step # 6, the optimum control parameters (Kpp, Kvp, Kvp,
The combination of Kvi) and the integrated value Ei (1) is stored in the control parameter storage unit 10, and the process ends.

With the above operation, the load change estimating unit 9 stores a row of data of the reference model of the deviation counter curve for each sampling cycle of position control within a certain fixed time from the start of the servo operation at the standard load. The control parameter storage unit 10 stores a deviation counter curve for each sampling cycle of position control within a certain fixed time from the start of servo operation when optimum control parameters are set for standard loads under various loads. A combination of the integrated value of the difference between the deviation counter curve and the reference model and the optimum control parameter for the load is stored as shown in Table 1.

[0030]

[Table 1]

In Table 1, Ei (0) is the case of the standard load, and the others show data for n kinds of loads.

Next, fuzzy inference in the control parameter auto tuning unit 11 will be described.

Control parameter auto-tuning unit 11
Calculates the overshoot amount OV and the settling time Ts from the deviation counter curve. First, an undershoot amount a as shown in FIG. 17 is obtained, and this is taken as an overshoot amount OV '. That is, 0V ′ = (a / b) × 10
0% Next, the time until the operation data falls within a fixed number of pulses (± 10 pulses in this embodiment) is obtained, and this time is settling time Ts'. And OV,
For the target values OVref and Tsref of Ts and the respective allowable ranges ΔOVref and ΔTsref, OV and Ts for fuzzy inference are determined based on the following arithmetic expression.

OV = {(OV′−OVref) / ΔOVref} +1 (where 0 ≦ OV ≦ (unit: percent) 20) Ts = (Ts′−Tsref) / ΔTsref (where 0 ≦ Ts ≦ (unit: ms) 50) Also, after the first undershoot (minimum point), undershoot (minimum point) exceeds the range of the above-mentioned constant number of pulses.
If there is, the time interval between the adjacent minimum points is obtained, and if the change in the time interval is within the set range and the minimum point exists five times or more, it is considered that there is oscillation. A value obtained by multiplying each of Kpp and Kvp by 0.9 at that time is set as the maximum value of the variable range of Kpp and Kvp, and the tuning is repeated within that range. If at least one of Kpp, Kvp, and α does not fall within the fixed range five times even after the tuning is repeated 10 times, α is lowered by a certain fixed value and tuning is performed again. After tuning Kpp, Kvp, and α, ΔK
pp, ΔKvp, Δα are ignored, and K is calculated based on ΔKvi.
Change vi. That is, Kvi + ΔKvi becomes the correction value. When ΔKvi becomes 0 or a negative value, K at that time
vi is stored, and when ΔKvi becomes 0 or a negative value appears five times, tuning is finished, and the minimum value among the stored five Kvi is multiplied by 0.9. The set value is set as the final Kvi.

Next, a target value for fuzzy inference is set.
That is, the target value and the allowable range of the overshoot amount and the settling time are set on the horizontal axis of the input membership function of the fuzzy inference, and the horizontal axis value of the vertex of the ZR triangle is set as shown in FIGS. A membership function with a target value and a permissible range at the base of the ZR triangle, and a preset membership function Figs. 5 to 8 and fuzzy rules Figs. 9 to 12.
On the basis of the overshoot amount OV and the settling time T
From the current value of s and how much the current acceleration α, the position amplifier proportional gain Kpp, the speed amplifier proportional gain Kvp, and the speed amplifier integral gain Kvi are corrected, the target overshoot amount OV and the settling time Ts can be approached. Reason.

First, the current values of the overshoot amount OV and settling time Ts calculated from the servo operation curve as described above are calculated. Depending on the OV and Ts, how much the current acceleration α, the position amplifier proportional gain Kpp, the speed amplifier proportional gain Kvp, and the speed amplifier integral gain Kvi / S (S indicates Laplace transform) can be changed to the target OV. Fuzzy inference is made as to whether it approaches Ts and Ts. This fuzzy inference method will be described below.

9 to 12, FIG. 9 shows the current K
This is a fuzzy rule for obtaining a correction amount ΔKpp for pp. 10, 11 and 12 show the current Kv, respectively.
It is a fuzzy rule for obtaining the correction amount ΔKvp for p, the correction amount ΔKvi for the current Kvi, and the correction amount Δα for the current α. For example, in the case of FIG. It means that. Each symbol is called a fuzzy label, NL is negatively large, NM is negatively medium,
NS is negatively small, ZR is almost zero, PS is positively small,
PM means positively medium and PL means positively large, and the input OV and Ts represent the states with respect to respective target values. For example, ZR means substantially target value. Become.

3 to 8, FIGS. 3 and 4 are graphs showing membership functions relating to OV and Ts calculated by the control parameter auto-tuning unit 11 as described above, FIGS. 5, 6, and 7, respectively. , FIG. 8 shows the output of the control parameter auto-tuning unit 11, ΔKpp,
It is a graph showing the membership function regarding ΔKvp, ΔKvi, and Δα. In this case, ΔKpp, ΔKvp,
ΔKvi and Δα are the current values Kpp, Kvp, and Kv, respectively.
i, α is a value obtained by multiplying the value on the horizontal axis of the membership function of FIGS. 5, 6, 7, and 8. For example, assuming that the inference result is 0.5 in ΔKpp, ΔKpp =
It becomes 0.5 × Kpp.

Control parameter auto-tuning unit 11
Is based on the membership functions shown in FIGS. 3 to 8 and the fuzzy rules shown in FIGS.
-Inference is performed by a fuzzy calculation method called the center of gravity method.

A method of calculating ΔKpp when OV = 1.6% and Ts = 8 (ms) will be described with reference to FIG.

First, OV = 1.6%, Ts = 8 (ms)
Finds how well fits each fuzzy label of each membership function. From FIG. 13, OV = 1.
6% has a goodness of fit of 0.4 for ZR and a goodness of fit of 0.6 for PS, and a goodness of fit of 0 for NS, PM and PL.
Is. Further, from FIG. 14, Ts = 8 (ms) has a goodness of fit of 0.3 for ZR, a goodness of fit of 0.7 for PS, and a goodness of fit of 0 for PM, PL, PL. .

Next, how much the combination of the input values of OV and Ts meets each rule is obtained. Figure 1
FIG. 6 shows only the fuzzy rules regarding ΔKpp in FIG. 9 that are established in the above case. The numerical values in FIG. 16 are the conformity values of each rule, and the conformity of each input is the smaller conformity by the MIN operation. Next, based on the goodness of fit of each rule shown in FIG. 16, as shown in FIG. 15, the fuzzy amount of the output ΔKpp is obtained by MAX calculation, and the center of gravity is obtained by the center of gravity method to determine the fixed output 0. 15 is obtained.
Then, the correction amount ΔKpp of Kpp is, as described above, Δ
Kpp = 0.15 × Kpp. Similarly, ΔKv
p, ΔKvi, and Δα are obtained.

Next, the procedure until the control parameter is driven to the optimum value (the target control characteristic is obtained) will be described.

First, the fuzzy inference result ΔKpp, Δ
Kpp, Kvp, and α are changed based on Kvp and Δα.
That is, Kpp + ΔKpp, Kvp + ΔKvp, α + Δα
Are new values. And ΔKpp, ΔKvp,
The tuning of Kpp, Kvp, and α is considered complete when the values of Δα are all within the respective fixed ranges for five consecutive times.

After the tuning of Kpp, Kvp, and α is completed, ΔKpp, ΔKvp, and Δα are ignored, and Kvi is changed based on ΔKvi. That is, Kvi + ΔKvi becomes a new value. Then, when ΔKvi becomes 0 or a negative value, Kvi at that time is stored, and when ΔKvi becomes 0 or a negative value is repeated 5 times, the tuning is finished and the stored value is stored. 5 times I saved Kvi
The value obtained by multiplying the minimum value of the above by 0.9 is set as the final Kvi.

In the above operation, the optimum control parameters for the standard load are always changed at the start of the servo operation. However, when the change in the load is not so large or the load applied to each axis of the robot. If it is known, each control parameter may be set to a constant value and the control parameter changing operation during acceleration may not be performed.

The automatic adjustment servo control device of the present invention is not limited to the above embodiment, but various modes are possible. For example,
In the embodiment, the number of repetitions is set to 5 times, 10 times, etc., but the number of times is not limited to 5 times, 10 times, and the number of times can be set according to the system. Further, although the control parameter gradual reduction coefficient is set to 0.9 in the embodiment, it is not limited to 0.9.

[0048]

According to the automatic adjustment servo control device of the present invention, even if the load changes during movement, the load size and the optimum control parameter are determined by defining the load size by the defining method of the present invention. Since the correspondence can be made on a one-to-one basis, the optimum control parameters can be easily read and used according to changes in the load, and the servo control using the optimum control parameters can always be performed.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of an automatic adjustment servo controller according to the present invention.

FIG. 2 is a plan view of an orthogonal robot.

FIG. 3 is a diagram of an input membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 4 is a diagram of an input membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 5 is a diagram of an output membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 6 is a diagram of an output membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 7 is a diagram of an output membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 8 is a diagram of an output membership function used for fuzzy inference according to an embodiment of the present invention.

FIG. 9 is a diagram of a fuzzy rule according to an embodiment of the present invention.

FIG. 10 is a diagram of a fuzzy rule according to an embodiment of the present invention.

FIG. 11 is a diagram of a fuzzy rule according to an embodiment of the present invention.

FIG. 12 is a diagram of a fuzzy rule according to an embodiment of the present invention.

FIG. 13 is an operation diagram of a fuzzy inference method according to an embodiment of the present invention.

FIG. 14 is an operation diagram of a fuzzy inference method according to an embodiment of the present invention.

FIG. 15 is an operation diagram of a fuzzy inference method according to an embodiment of the present invention.

FIG. 16 is an operation diagram of a fuzzy inference method according to an embodiment of the present invention.

FIG. 17 is a diagram of a servo control curve according to an embodiment of the present invention.

FIG. 18 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 19 is a block diagram of a conventional example.

[Explanation of symbols]

 1 Command Control Section 2 Deviation Counter 3 Position Amplifier 4 Speed Amplifier 5 Drive Circuit 6 Motor 7 Position Detector 8 Speed Detection Section 9 Load Change Estimation Section 10 Control Parameter Storage Section 11 Control Parameter Auto Tuning Section 12 Control Mechanism

Continuation of front page (51) Int.Cl. ⁵ Identification code Office reference number FI Technical display location G05B 13/02 N 9131-3H 19/403 V 9064-3H

Claims (2)

[Claims] 1. A command control unit that outputs a position command value in order to match the movement of a servo controlled object to a target position at each time by controlling a motor, and detects a difference between the position command value and an actual position. In a servo control device having a deviation counter that outputs a deviation counter curve and a position control loop and a speed control loop that operate based on control parameters, a load change estimation unit that defines the magnitude of the load from the deviation counter curve. , A control parameter storage unit that stores preset magnitudes of various loads and optimum control parameters corresponding thereto, and controls the entire control device, and at the start of servo operation, preset optimum parameters at standard load Is used, and the deviation change curve at the start of the servo operation and The difference between the deviation counter curve and the reference model when the optimum parameters are used with a preset standard load is integrated, and the actual load size is defined from the result, and at the end of the certain time, the above defined An automatic control system comprising: a control mechanism that reads out an optimum control parameter corresponding to the actual load size from the control parameter storage unit based on the actual load size and controls using the optimum control parameter. Adjusting servo controller. 2. A control parameter auto-tuning unit that calculates an overshoot amount and a settling time from a deviation counter curve and adjusts a control parameter by fuzzy reasoning from the calculation result, and the entire control device by controlling the control parameter. When the auto tuning unit calculates the optimum control parameter under standard load, stores the optimum control parameter in the control parameter storage unit, performs servo operation using the optimum control parameter under the standard load, and starts servo operation. From this, the deviation counter curve for a fixed time is obtained as the reference model of the deviation counter curve,
This reference model is stored in the control parameter storage unit, only the load is sequentially changed to various loads, and in the case of these various loads, the load change estimation unit is set to the load change estimation unit for a fixed time from the servo operation start time. An integral value of the difference between the deviation counter curve and the reference model of the deviation counter curve is obtained, and at the same time, the control parameter auto-tuning unit is caused to calculate the optimum control parameter for the various loads, and at the end of the certain time, the The automatic adjustment servo control device according to claim 1, further comprising a control mechanism that stores a combination of the integrated value of the difference and the optimum control parameter in the control parameter storage unit.