JP2006331013A - Servo controller and control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a servo controller and a servo control method for improving control performance by estimating status quantity n sampling ahead from a torque command and a position detection value by a small memory capacity and a simple formula. <P>SOLUTION: This servo controller is provided with a position control unit, a speed control unit and a current control unit, and configured to perform a control arithmetic operation for operating a control object as commanded from a command and a detection value to be output by a detector added to the control object for generating current I. This servo controller is provided with an n sampling ahead estimation unit, and the n sampling ahead estimation unit is characterized to perform the arithmetic operation of the following expression. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a servo control device and a control method that can improve control characteristics by improving delay of a control system.

The conventional servo control method improves the delay by predicting the n-th state quantity with a predictor (see, for example, Patent Document 1).
FIG. 9 is a block diagram of a conventional motor speed control device. FIG. 10 is a block diagram of the observer shown in FIG. FIG. 11 is a block diagram of the predictor shown in FIG. In FIG. 9, reference numeral 86 denotes a control arithmetic unit which inputs a speed command and a speed feedback signal vfb and outputs a torque command u. Reference numeral 87 includes a torque controller, a motor, and a position detector. At the current time i, a torque command u (i) is input, and a motor position detection value before K · Ts (K ≧ 0, Ts: sampling period). y (i-K) is output. 82 is a velocity v (i−K) from the position y (i−K) by an operation such as v (i−K) = Δy (i−K) / Ts (Δ represents an increment value during the sampling period Ts). ). Reference numeral 85 denotes an observer, which obtains an estimated speed value v ^ from the torque command u, speed v, motor and disturbance dynamic characteristic model, and outputs it to the predictor 81 and the memory 83. Reference numeral 81 denotes a predictor for obtaining a predicted speed value v ^ * (i + m) m = −K + 1,..., M up to M sampling destinations from a motor dynamic characteristic model, a torque command u, and a speed estimated value v ^. Reference numeral 83 denotes a memory for storing a speed estimated value v ^ (im) m = K,..., M ′ from before M ′ sampling to time i−K. 84 is an arithmetic unit,

  As a result, the speed feedback signal vfb (i) is obtained and output to the control calculator 86. The weights Wm and W′m are set to be 1 as a whole. By setting M ≧ M ′ for M and M ′, a speed feedback signal vfb (i) having no phase delay can be obtained.

  Next, a specific operation of the observation device 85 will be described. Now, assuming state variable x, control input u, and control output y, state equation and output equation

  If there is a stable control system that yields, the unobservable state variable x is

It is widely known that x ^ can be obtained by configuring the control system of the above and making the matrix A-KC a stable matrix. Therefore, when the speed v, the disturbance torque ud are state variables, the torque command u is a control input, the transfer function model from the torque command to the speed is 1 / Js, and the disturbance model is a constant disturbance, the equation (2a)

Equation (3) is obtained, and when the control output is the velocity v, the velocity estimation value v ^ and the disturbance estimation u ^ d are obtained from the equation (3).

  Configure the control system of

  Can be obtained by determining k1 and k2 such that becomes a stable matrix. The stability matrix is the determinant

  Is established. For example, if k1 and k2 are determined so that the equation (6) becomes a multiple root η, k1 = 2η and k2 = η2 are obtained. u ^ d can be obtained. FIG. 10 shows an example of the control system of the formula (5) by the observer 5. From the configuration of FIG. 10, the speed estimated value v ^ has a characteristic that the speed v is filtered with characteristics determined by k1 and k2. Therefore, the resolution of the position detector is low, and the accuracy of the speed v by the computing unit 82 by quantization or the like. Even if it is bad, it has a feature that a smooth speed can be obtained.

  Next, the specific operation of the predictor 81 is shown in FIG. In FIG. 11, 81 is a predictor, 92 is a memory for storing prediction coefficients Amn and Bmn, and 93 and 94 are memories for storing past torque commands u and speed estimation values v ^ up to the present. 91 is an arithmetic unit,

  The predicted speed value v ^ * (i + m) is calculated and output by the following calculation. Here, the prediction coefficients Amn and Bmn will be described. Now, the transfer function model from the torque command u to the estimated speed v ^ is

  Is obtained in the discrete time system, the input / output model is as follows.

  At time i, since the estimated speed value v ^ (in) (n ≧ K) up to time i−K is obtained, the estimated speed value thereafter is

The predicted speed value v ^ * (i + m) is expressed by Equation (7), and the prediction coefficients Amn and Bmn are set so that the future torque command is u (j) = 0 (j> i).

Given in. If u (j) = u (i) (j> i), Bmo in the equation (11b) is as follows.

As described above, in the conventional servo control device, the estimated speed value is calculated by the observer, the predicted speed value is calculated by the predictor, and the M ′ past estimated speed value stored in the memory and the M estimated speed value stored in the memory are calculated. A procedure has been taken in which a speed feedback value is calculated from the previous speed predicted value and used in a control computing unit.
Japanese Patent Laying-Open No. 2002-354861 (page 5-7, FIG. 1-3)

In the conventional servo control method, the estimated speed value is calculated by the observer, the estimated speed value is calculated by the predictor, and the M ′ past estimated speed value stored in the memory and the predicted M speed value stored in the memory are calculated. Since the procedure of calculating the speed feedback value from the value and using it in the control arithmetic unit is taken, a complicated and enormous amount of calculation is required. Further, since it is necessary to always store at least M + M ′ variables in the memory, there is a problem that a huge memory area is required. As a result, there is a problem that the conventional method cannot be realized with a CPU having a low calculation processing speed and an inexpensive controller configured with a small memory area.
The present invention has been made in view of such problems, and can estimate a state quantity of n sampling destinations from a torque command and a position detection value with a small memory capacity and a simple expression, and reduce a delay in speed detection. Speed controller response can be made faster, disturbances can be compensated without delay in the control system, and not only an integer multiple of the sampling period, but also a state quantity estimated value whose phase is advanced by that time An object of the present invention is to provide a servo controller and a control method.

In order to solve the above problem, the present invention is as follows.
According to the first aspect of the present invention, a position controller that generates a speed command from the position command and a detection value generated by a detector added to the control target, and a torque command tref from the speed command and the speed detection value. A servo comprising a speed controller, a current controller that generates the current I from the torque command and drives the controlled object, and an n sampling estimator that estimates the state quantity of the controlled object from the detected value and the torque command In the servo control method of the control device, the current value obtained by dividing the torque command tref by the inertial value Jn, that is, the k-th input uref (k) and the current position detection value x (k) are input. , One sampling destination, that is, the (k + 1) th position estimated value xh (k + 1) and the (k + 1) th speed estimated value vh (k + 1) and the (k + 1) th outside Estimated value dh (k + 1) is estimated, and with respect to an arbitrary n, the following n-th position, that is, k + nth position estimated value xh (k + n) and k + nth speed estimated value vh (k + n) and k + nth The disturbance estimated value dh (k + n) is estimated, and the estimated n number of state quantities are used in the control calculation.

Where Ts: control sampling period, L1 and L2 and L3: observer gain,! : Indicates factorial calculation.
According to the second aspect of the present invention, at least the speed controller that generates the torque command tref from the speed command and the detection value generated by the detector added to the control target, and the current I from the torque command are generated. In a servo control method of a servo control device comprising: a current controller that drives the control object; and an n-sampling estimator that estimates the state quantity of the control object from the detection value and the torque command, the torque command tref is inertial nominal This time, i.e., the kth input uref (k) divided by the value Jn and the current speed detection value v (k) are input, and the following equation, where n is 1, 1 sampling destination, i.e., k + 1th The estimated speed value vh (k + 1) and the (k + 1) th disturbance estimated value dh (k + 1) are estimated, and again n times ahead of an arbitrary n, that is, k by the following equation: The n-th speed estimated value vh (k + n) and the k + n-th disturbance estimated value dh (k + n) are estimated, and the estimated n number of state quantities are used in the control calculation. It is.

Here, Ts: control sampling period, L1 and L2: observer gain,! : Indicates factorial calculation.
According to a third aspect of the present invention, in the servo control method according to the first or second aspect, the input uref (k) is obtained by multiplying the current value I (k) by the torque multiplier Kt and dividing by the inertial nominal value Jn. A signal is used.
According to a fourth aspect of the present invention, in the servo control method according to the first or second aspect, the speed estimated value vh (k + n) is a speed feedback value.
According to a fifth aspect of the present invention, in the servo control method according to the first or second aspect of the invention, the disturbance estimated value dh (k + n) is multiplied by a gain, a filter process is performed, and a torque command tr (k) is obtained. It is characterized by adding.
According to a sixth aspect of the present invention, in the servo control method according to the first or second aspect, in the process of using the estimated n-th state quantity in the control calculation, a plurality of n-th estimated values are used. Each is obtained by calculation, multiplied by a weight such that the total is 1, and the sum of all of them is used.
The invention according to claim 7 has a position controller, a speed controller, and a current controller, and controls the control object from the detected value output from the command and the detector added to the control object. In the servo control device that performs the control operation to generate the current I and generates the current I, the servo control device includes an n sampling destination estimator, and the n sampling destination estimator performs the calculation of the following equation: .

Where Ts: control sampling period, L1 and L2 and L3: observer gain,! : Indicates factorial calculation.
The invention according to claim 8 has a position controller, a speed controller, and a current controller, and controls the control object from the detected value output from the command and the detector added to the control object. In the servo control device that performs the control operation to generate the current I to generate the current I, the servo control device includes an n sampling destination estimator, and the n sampling destination estimator performs the operation of the following expression.

Here, Ts: control sampling period, L1 and L2: observer gain,! : Indicates factorial calculation.
The invention according to claim 9 includes a speed controller and a current controller, and controls the control object to operate according to the command from the command and the detection value output from the detector added to the control object. The servo control device that performs the calculation to generate the current I includes an n sampling destination estimator, and the n sampling destination estimator performs the calculation of the following expression.

Here, Ts: control sampling period, L1 and L2: observer gain,! : Calculation of factorial, v (k): speed detection value, vh (k): kth speed, dh (k): kth disturbance estimated value, vh (k + n): k + nth speed, dh (k + n) ): K + n-th disturbance estimated value, uref (k): represents a value obtained by dividing the torque command tref by the inertial nominal value Jn.

According to the first and seventh aspects of the present invention, the amount of memory used for simple calculation is small and the n-state ahead state quantity can be estimated. Deterioration of control performance can be improved.
Further, according to the invention described in claim 2 and claim 8, the state quantity n ahead can be estimated with a smaller amount of calculation, and the control performance due to delay can be obtained by using the estimated value ahead n. Deterioration can be improved.
According to the invention described in claim 3, by using the current value instead of the torque command value, it is possible to estimate the n-th state quantity more accurately and use the n-th ahead estimated value in the control. Thus, deterioration of control performance due to delay can be improved.
According to the invention described in claim 4, by using the n-th ahead estimated speed value as the speed feedback value, the speed detection delay can be improved and the speed controller response can be increased. The deterioration of performance can be improved.
Further, according to the fifth aspect of the present invention, the n-th ahead estimated disturbance value can be calculated, so that the control system can compensate for the disturbance without delay, and as a result, high-performance control can be realized.
Further, according to the invention described in claim 6, since it is possible to obtain the state quantity estimated value whose phase is advanced not only by an integral multiple of the sampling period but also by the time between them, the delay can be compensated more strictly, and as a result Control with good performance can be realized.

  Hereinafter, specific examples of the method of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing the configuration of a servo control apparatus that implements the method of the present invention. In the figure, reference numeral 1 denotes a control object, which is composed of an electric motor and a load machine, for example. 2 represents a detector and detects the position of the motor. Reference numeral 100 denotes the entire servo control device.

Next, the internal configuration of the servo control device will be described.
Reference numeral 3 denotes a position controller, which controls the current position, that is, the k-th position command ref (k) and the current position detection value x (k) to output a speed command vr (k). Reference numeral 4 denotes a speed controller, which inputs a speed command value vr (k) and a speed feedback value and outputs a torque command value tref (k). The current controller No. 5 receives the torque command value tref (k) and controls the electric motor to generate a torque according to the torque command value to generate a current value I (k). Normally, the position controller performs proportional control, and the speed controller and the current controller perform proportional-integral control. Reference numeral 7 denotes an n sampling destination estimator, which estimates an n-th ahead state quantity based on the torque command value tref (k) and the position detection value x (k).

In this embodiment, the response of the speed controller can be improved by using the estimated speed value vh (k + n) among the estimated state quantities as the speed feedback value.
As processing inside the n sampling destination estimator, only the following equation (12) is calculated.

Where Ts: control sampling period, L1 and L2 and L3: observer gain,! : Factorial calculation, xh (k): kth position, vh (k): kth speed, dh (k): kth disturbance estimate, xh (k + n): k + nth position, vh ( k + n): k + n-th velocity, dh (k + n): k + n-th disturbance estimated value.
The observer gains L1 to L3 may be appropriately set so that the observer poles are at desired positions. For example, calculation may be performed so that the pole when n = 1 is a triple root of 200 Hz.
Further, uref (k) may be obtained by the equation (13) using the inertial nominal value Jn.
uref (k) = tref (k) / Jn (13)
When actually performing the calculation, first, using equation (12), the k + 1-th state quantity is calculated with n = 1. This is used as xh (k), vh (k), and dh (k) in the next calculation. Then, the estimated value for the nth time (for example, if n = 3 is three ahead) may be obtained by calculation.

  By simply using the calculated speed estimation value vh (k + n) as a speed feedback value, it is possible to improve the adverse effects caused by delays such as speed detection delay, calculation delay and communication delay, and increase the speed controller gain. Can be realized.

Also, instead of using only one n, a plurality of n-th estimated values are obtained by calculation, respectively, multiplied by a weight such that the total is 1, and a state quantity obtained by summing all of them is obtained. By using it, not only an integer multiple of the sampling period, but also the state quantity estimate with the phase advanced by that time can be obtained, so the delay can be compensated more strictly, and as a result, good performance control can be realized. Can do. For example, when it is desired to advance the estimated speed value by 1.5 in the sampling cycle, the estimated values of n = 1 and n = 2 may be multiplied by a weight of 0.5 as shown in Expression (14).
vh (k + 1.5) = 0.5 · vh (k + 1) + 0.5 · vh (k + 2) (14)

FIG. 2 is a block diagram showing a second embodiment. The difference from the first embodiment is only that the input of 7 n sampling destination estimators is I (k) instead of tref (k).
In this case, uref (k) may be calculated as shown in Expression (15) using a torque constant Kt for converting current and torque.
uref (k) = I (k) · Kt / Jn (15)

  FIG. 3 is a block diagram showing a third embodiment. The difference from the first embodiment is that the n-th ahead estimated estimated value dh (k + n) is not used in the speed control, but the n-th estimated destination estimated value dh (k + n) is used in the control. It is. Specifically, as shown in FIG. 8, the signal tc (k) obtained by multiplying dh (k + n) by the gain G and performing, for example, the first-order low-pass filter processing and the speed controller 4 inside the gain / filter processor 8. The output tr (k) may be added to obtain the torque command value tref (k). Here, any filter may be used, and a high-pass filter or a band-pass filter may be used. The gain G can use both positive and negative values.

  FIG. 4 is a block diagram showing a fourth embodiment. The only difference from the third embodiment is that the input of the n sampling destination estimator is I (k) instead of tref (k). Therefore, if Expression (15) is used, it can be realized by the same processing as in the third embodiment.

  FIG. 5 is a block diagram showing a fifth embodiment. The difference from the first embodiment is that the input of the n sampling destination estimator is not the x (k) but the detected speed value v (k). In this case, the formula for calculating the n-th state quantity is also calculated using Formula (16) instead of Formula (12).

Here, the speed detection value v (k) may be obtained by calculation using the position detection value x (k). In this case, the speed detection value v (k) may be calculated using the difference as shown in Expression (17).
v (k) = {x (k) -x (k-1)} / Ts (17)
When the n-th estimated speed value vh (k + n) is obtained, the speed controller may use it as a speed feedback value as in the first embodiment.

  FIG. 6 is a block diagram showing a sixth embodiment. The only difference from the fifth embodiment is that the input of the n sampling destination estimator is I (k) instead of tref (k). Therefore, if Expression (15) is used, it can be realized by the same processing as in the fifth embodiment.

  FIG. 7 is a block diagram showing a seventh embodiment. In the present embodiment, the input of the n sampling destination estimator is not the x (k) but the speed detection value v (k) in the configuration of the fourth embodiment. Therefore, it calculates using Formula (16). Since the present embodiment is a combination of the fourth embodiment and the fifth embodiment, detailed description thereof is omitted.

  FIG. 8 is a block diagram showing an eighth embodiment. The only difference from the seventh embodiment is that the input of the n sampling destination estimator is I (k) instead of tref (k). Therefore, if Expression (15) and Expression (16) are used, the subsequent processing can be realized by exactly the same processing as in the seventh embodiment.

As described above, the present system can be used in various configurations.
In the present embodiment, the configuration for performing the position control has been described. Naturally, the effect can be obtained by the same processing in the case of the speed control.

  Since it is possible to estimate the state quantity ahead n by simple calculations, it is possible to achieve optimal control for any axis by adjusting n according to the delay of each axis, such as when communicating with multiple axis controllers. It can be applied to servo control applications such as multi-axis position, speed, and trajectory, such as robots, machine tools, and semiconductor manufacturing equipment.

The block diagram which shows the structure of the servo control apparatus to which the method of this invention is applied. The block diagram explaining the structure of the 2nd Example of this invention. The block diagram explaining the structure of the 3rd Example of this invention The block diagram explaining the structure of the 4th Example of this invention. The block diagram explaining the structure of 5th Example of this invention The block diagram explaining the structure of the 6th Example of this invention. The block diagram explaining the structure of the 7th Example of this invention The block diagram explaining the structure of the 8th Example of this invention. Block diagram showing the configuration of a conventional servo controller Block diagram showing the processing inside the conventional servo control method observer Block diagram showing the internal processing of the predictor of the conventional servo control method

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Control object 2 Detector 3 Position controller 4 Speed controller 5 Current controller 6 Differentiator 7 n sampling destination estimator 8 Gain filter processor 81 Predictor 82 Calculator 83 Memory 84 Calculator 85 Observer 86 Control calculation Device 87 Motor 100 Servo control device 111 Calculator 112 Memory 113 Memory 114 Memory

Claims (9)

A position controller that generates a speed command from a position command and a detection value generated by a detector added to the object to be controlled; a speed controller that generates a torque command tref from the speed command and the speed detection value; In a servo control method of a servo control device comprising: a current controller that generates a current I to drive the controlled object; and an n sampling estimator that estimates a state quantity of the controlled object from the detected value and the torque command.
The current command obtained by dividing the torque command tref by the inertial nominal value Jn, that is, the k-th input uref (k) and the current position detection value x (k) are input.
In the following equation, when n is 1, 1 sampling destination, that is, (k + 1) th position estimated value xh (k + 1), (k + 1) th velocity estimated value vh (k + 1), and (k + 1) th disturbance estimated value dh (k + 1) Estimate
Again, for any n,
n-th, ie, k + n-th position estimation value xh (k + n), k + n-th velocity estimation value vh (k + n) and k + n-th disturbance estimation value dh (k + n);
A control method for a servo control device, wherein the estimated n state quantities are used in a control calculation.

Where Ts: control sampling period, L1 and L2 and L3: observer gain,! : Indicates factorial calculation. A speed controller that generates a torque command tref from at least a speed command and a detection value generated by a detector added to the control target; a current controller that generates a current I from the torque command and drives the control target; In the servo control method of the servo control device comprising an n sampling estimator for estimating the state quantity of the controlled object from the detected value and the torque command,
The present time obtained by dividing the torque command tref by the inertial value Jn, that is, the kth input uref (k) and the current speed detection value v (k) are input.
Estimate one sampling destination, that is, the (k + 1) th estimated velocity value vh (k + 1) and the (k + 1) th disturbance estimated value dh (k + 1) by the equation when n in the following equation is 1.
Again, the following equation is used to estimate the n-th or k + n-th estimated velocity value vh (k + n) and the k + n-th disturbance estimated value dh (k + n) with respect to any n.
A control method for a servo control device, wherein the estimated n state quantities are used in a control calculation.

Here, Ts: control sampling period, L1 and L2: observer gain,! : Indicates factorial calculation.   3. The control method for a servo control device according to claim 1, wherein the input uref (k) uses a signal obtained by multiplying the current value I (k) by a torque multiplier Kt and dividing it by an inertial value Jn. 3. The servo control device control method according to claim 1, wherein the speed estimation value vh (k + n) is a speed feedback value.   The servo control device control method according to claim 1, wherein the disturbance estimated value dh (k + n) is multiplied by a gain, subjected to a filtering process, and added to a torque command tr (k).   In the process of using the estimated n number of state quantities in the control calculation, a plurality of n number of estimated values are obtained by calculation, respectively, and multiplied by a weight such that the sum is 1, respectively, The control method of the servo control device according to claim 1, wherein the control device is used. A position controller, a speed controller, and a current controller are provided, and a control calculation is performed to operate the control target as instructed from the command and the detection value output from the detector added to the control target. In the servo control device that generates
n sampling destination estimators,
The n sampling destination estimator performs a calculation of the following equation:

Where Ts: control sampling period, L1 and L2 and L3: observer gain,! : Calculation of factorial, x (k): position detection value, xh (k): kth position, vh (k): kth speed, dh (k): kth disturbance estimated value, xh (k + n) ): K + nth position, vh (k + n): k + nth speed, dh (k + n): k + nth estimated disturbance value, uref (k): value obtained by dividing the torque command tref by the inertial nominal value Jn. A position controller, a speed controller, and a current controller are provided, and a control calculation is performed to operate the control target as instructed from the command and the detection value output from the detector added to the control target. In the servo control device that generates
n sampling destination estimators,
The n sampling destination estimator performs a calculation of the following formula:

Here, Ts: control sampling period, L1 and L2: observer gain,! : Calculation of factorial, v (k): speed detection value, vh (k): kth speed, dh (k): kth disturbance estimated value, vh (k + n): k + nth speed, dh ( k + n): k + nth estimated disturbance value, uref (k): a value obtained by dividing the torque command tref by the inertial nominal value Jn. Servo control that has a speed controller and a current controller, and generates a current I by performing a control operation to operate the controlled object according to the command from the command and the detection value output by the detector added to the controlled object In the device
n sampling destination estimators,
The n sampling destination estimator performs a calculation of the following formula:

Here, Ts: control sampling period, L1 and L2: observer gain,! : Calculation of factorial, v (k): speed detection value, vh (k): kth speed, dh (k): kth disturbance estimated value, vh (k + n): k + nth speed, dh ( k + n): k + nth estimated disturbance value, uref (k): a value obtained by dividing the torque command tref by the inertial nominal value Jn.