JP6979330B2 - Feedback control method and motor control device - Google Patents

Description

The present invention relates to a feedback control method and a motor control device including the control method.

In recent years, in the FA field, more and more high-speed and high-precision control of motors is required to improve productivity.

When feedback-controlling the motor, the control gain should be increased in order to suppress disturbance and make the controlled variable follow the target value at high speed and with high accuracy. However, if there is a delay element (for example, calculation delay of a low-pass filter or digital control device) in the feedback loop, the upper limit of the control gain setting of the feedback control system is restricted due to this, and high-speed and highly accurate target value tracking can be performed. It is generally known to be an obstacle.

As a design method of a delay compensator that can compensate for the delay element existing in the closed loop of the feedback control system and suppress the step disturbance applied to the input end of the control target without steady deviation even when the control target has a pole at the origin. Patent Document 1 has been proposed.

In Patent Document 1, as shown in FIG. 6, in the delay compensator 62 (N = 1, which matches the conventional Smith method) in which the filter 61 is added to the Smith method which is the prior art, the control target has a pole at the origin. The design method of the filter 61 is shown, which has the characteristics that the step disturbance applied to the input end of the controlled object can be suppressed without steady deviation even if the filter 61 is provided, and the physical meaning of the design parameters of the filter 61 is easy to understand. .. By adopting the filter 61 having various configurations designed by this design method for the delay compensator 62, various delay compensators 62 having the above-mentioned characteristics can be designed.

Japanese Patent Application No. 2017-054595

In Patent Document 1, the configurations of the following equations (1) and (2) are given as design examples of the filter 61 shown in FIG.

However, Pm is a controlled plant model 14, Cb (θ1) has θ1 as an adjustment parameter, and a feedback controller 16 capable of suppressing a deviation between a target value and a controlled object response is included in the controlled object 12. The model value exp (−τm · s) of the sum of the delay time τf and the delay time τb included in the feedback delay element 13 is a nominal delay element model 15 due to the delay time τm.

Further, Ca is an arbitrary feedback controller that functions effectively with respect to the controlled object, and when Ca has the same structure as the feedback controller Cb, Ca is defined as in the equation (3).

However, the filters 61 represented by the equations (1) and (2) include exp (-τm · s) in the denominator, and of course, when exp (−τm · s) is calculated exactly, this is used. Even when approximating with a low-order transfer function using the Padé approximation method or the like, the order of the transfer function of the filter represented by the equations (1) and (2) tends to be high.

Therefore, for example, when the filter 61 of the delay compensator 62 shown in FIG. 6 designed in Patent Document 1 represented by the equations (1) and (2) is mounted on hardware such as a digital arithmetic unit, the hardware is used. There is a problem that the calculation cost required for the above filtering process becomes high.

The present invention has been made in view of such a problem, and a delay compensator 62 including, for example, the filters 61 represented by the equations (1) and (2), designed by the design method of Patent Document 1, is provided. It is an object of the present invention to provide a control method suitable for mounting, which can reduce the calculation cost of the hardware applied to the filter 61 of the delay compensator 62 when mounted on hardware such as a digital arithmetic unit, and a control device provided with the control method. And.

The present invention is a feedback control method including a delay compensator 1 composed of a model to be controlled, an arbitrary filter 1, and an arbitrary feedback controller 1, to give an example in view of the above background technology and problems. Therefore, the model to be controlled consists of a nominal plant model that simulates the dynamics of the controlled object and a nominal delay model that simulates the delay element contained in the closed loop of the feedback control system, and the delay compensator 1 controls. The operation amount output by the feedback controller 2 that performs feedback control to the target and the output signal of the control target are used as input signals, and the operation amount and the output of an arbitrary feedback controller 1 included in the delay compensator 1 are added or subtracted. The output signal of the nominal plant model for the signal added by the device is used as the ideal feedback signal, and the control target model for the signal obtained by adding the operation amount and the output of any feedback controller 1 included in the delay compensator 1 by the adder / subtractor. The signal obtained by subtracting the output signal and the output signal to be controlled by an adder / subtractor is used as a model error signal, the model error signal is used as an input of an arbitrary feedback controller 1, and the model error signal is used as an arbitrary filter 1. The output signal is a signal obtained by adding the processed signal and the ideal feedback signal by the addition / subtractor, and the feedback controller 2 calculates the deviation between the output signal of the delay compensator 1 and the target value signal by the addition / subtractor. Then, feedback compensation is performed for the controlled object based on the deviation.

According to the present invention, it is possible to provide a feedback control method suitable for mounting, which can reduce the calculation cost in hardware mounting, and a motor control device including the feedback control method.

It is a block diagram (A (s) = 1) of the model follow-up type control system in Example 1. FIG. It is a block diagram of the model follow-up type control system in Example 1. FIG. It is a block diagram of the feedback control system including the delay compensator in Example 1. FIG. It is a block diagram (A (s) = 1) of the feedback control system including the delay compensator in Example 1. FIG. It is a block diagram (A (s) = γ) of the feedback control system including the delay compensator in Example 1. FIG. It is a block diagram of the feedback control system including the delay compensator by the prior art. It is a block diagram of the speed control system of an AC servomotor. It is a block diagram of the speed feedback control system including the delay compensator in the prior art. It is a block diagram of the speed feedback control system including the delay compensator in Example 2. FIG.

Hereinafter, examples of the present invention will be described with reference to the drawings. In each figure, the same numbers are assigned to the components having a common function, and the description thereof will be omitted. Hereinafter, "feedback" is abbreviated as "FB". For example, "Feedback controller is abbreviated as" FB controller ". Further, "feed forward" is abbreviated as "FF". For example, "feedforward controller is abbreviated as" FF controller ".

In the motor control method according to the present embodiment and the motor control device provided with the motor control system, as shown in FIG. 3, the FB control system is used by the delay compensator 2 and the FB controller 16 for the control target 12 including the delay. Is configured.

In FIG. 3, CA and A (s) are arbitrary FB controllers 3 and filters 1, respectively, which function effectively for the controlled object. In this embodiment, these are used to derive compensation performance equivalent to the delay compensator 62 shown in FIG. 6 having the filter 61 (for example, the above-mentioned equations (1) and (2)) designed in Patent Document 1. It is preferably configured.

The delay compensator 2 is composed of an FB controller 3, a filter 1, and a model to be controlled. In this embodiment, the model to be controlled is composed of a nominal plant model 14 and a nominal delay element model 15, which simulates the dynamic characteristics of the control target 12 and the FB delay element 13 which are the actual control targets. Is. That is, the model to be controlled consists of a nominal plant model that simulates the dynamics of the controlled object and a nominal delay model that simulates the delay element contained in the closed loop of the feedback control system.

When elements that cause delays such as a filter and a minor loop control system are included in the closed loop system, the model to be controlled may include a nominal model of those delay elements.

Hereinafter, the transmission characteristic in which the control target 12 and the FB delay element 13 are connected in series may be simply referred to as a control target.

The delay compensator 2 has an operation amount output by the FB controller 16 and an output of a controlled object for the operation amount output by the FB controller 16 as inputs, and outputs an FB signal with respect to the target value r.

The operation of the delay compensator 2 will be described with reference to FIG. The delay compensator 2 is an operation in which the FB controller 16 outputs the response of the model to be controlled to the signal obtained by adding the output of the arbitrary FB controller 3 and the operation amount output by the FB controller 16 by the adder / subtractor 4. The model error signal is calculated by subtracting from the response output of the controlled object to the quantity by the adder / subtractor 39.

Further, the delay compensator 2 calculates the response of the plant nominal model 14 to the signal obtained by adding the output of the arbitrary FB controller 3 and the operation amount output by the FB controller 16 by the adder / subtractor 4 as a predicted FB signal. The signal obtained by processing the model error signal by the filter 1 and the predicted FB signal are added by the adder / subtractor 35 and output as an FB signal with respect to the target value r.

The FB controller 16 receives a deviation signal obtained by subtracting the FB signal output by the delay compensator 2 from the target value r by the adder / subtractor 37 as an input, and generates an operation amount for the controlled object.

In the FB control system including the delay compensator 2 shown in FIG. 3, the adder / subtractor 18 is located at the input end of the controlled object, and it is assumed that the disturbance d is applied to the input end of the controlled object.

When the delay compensator is configured by the generally known Smith method (in the case of the delay compensator 62 where N = 1 in FIG. 6), when the controlled object has a pole at the origin, a step to be applied to the input end of the controlled object. There was a problem that the disturbance could not be suppressed without steady deviation. On the other hand, in the case of the delay compensator 62 designed in Patent Document 1, for example, the delay compensator 62 using the filters N of the equations (1) and (2), the step disturbance applied to the input end to be controlled is suppressed without steady deviation. be able to.

The delay compensator 2 shown in FIG. 3 in this embodiment has compensation performance equivalent to that of the delay compensator 62 designed in Patent Document 1, for example, which employs the filters N of the equations (1) and (2). It is designed.

For this purpose, the CA and A (s) of the delay compensator 2 are designed, for example, as the following equations (4) and (5).

However, Ca indicates an arbitrary feedback controller that functions effectively with respect to the controlled object, which was used in designing the filter 61 of the delay compensator 62 shown in FIG. 6 in Patent Document 1.

The delay compensator designed as in the equations (4) and (5) corresponds to the delay compensator 42 in FIG. 4, and the input / output characteristics of the delay compensator 42 are shown by the equation (1) from a simple calculation. The input / output characteristics are equivalent to the delay compensator 62 of FIG. 6 including the filter N. This will be described below.

The input / output characteristics of the delay compensator 62 in FIG. 6 can be written by the following equation.

On the other hand, the input / output characteristics of the delay compensator 42 in FIG. 3 can be written as the equation (9) by arranging the following equations (7) and (8) derived from FIG. 3 with respect to the input and output.

Equation (9) becomes equal to equation (6) by making equations (4) and (5). Therefore, with respect to the delay compensator 2 of FIG. 3, the delay compensator 42 designed as shown in the equations (4) and (5) includes the filter N represented by the equation (1) in the delay compensator of FIG. It can be seen that the input / output characteristics are equivalent to those of the device 62.

Further, the CA and A (s) of the delay compensator 2 are designed, for example, as the following equations (10) and (11).

The delay compensator 2 designed as shown in the equations (10) and (11) includes the filter N represented by the equation (2) by the same equation expansion as the equations (6) to (9). It can be confirmed that the input / output characteristics are equivalent to those of the delay compensator 62.

Equations (4) to (5) and equations (10) to (11) all include the transfer characteristic Pm of the nominal plant model 14 included in the equations (1) and (2), and the nominal delay element model 15. It can be seen that the transfer function is simple and does not include the transfer characteristic exp (-τm · s).

The reason why the transfer functions of CA and A (s) are simplified as compared with the equations (1) and (2) is that Pm and exp (-) included in the equations (1) and (2). This is because the delay compensator 2 is configured as a block configuration so that the calculation of τm · s) can be shared with the calculation of the nominal plant model 14 and the nominal delay element model 15 in the figure.

Next, the characteristics of the delay compensator 2 shown in FIG. 3 will be described with reference to FIGS. 1 and 2. The command value response r → y and the disturbance response d → y in FIGS. 1 and 2 are equivalent to the command value response r → y and the disturbance response d → y shown in FIGS. 4 and 3, respectively, that is, FIG. FIG. 2 is a control system whose response is equivalent to that of FIGS. 4 and 3, respectively. Further, FIG. 1 is a case where A (s) = 1 in FIG.

First, the characteristics of the delay compensator 2 will be described with reference to FIG. In FIG. 1, the transmission characteristic of rs → y is given by the normative model type real FB control system 7. Hereinafter, for the sake of simplification of the explanation, it is assumed that there is no disturbance and d = 0. However, even if it was d ≠ 0, the following description of the particular note is established.

The model FB control system 8 is configured as shown in FIG. 1 by using the nominal plant model 14, the nominal delay element model 15, and the FB controller CA3. In the model FB control system 8, y corresponds to a command, and u can be regarded as an FF operation amount in response rs → ym. Therefore, the model FB control system 8 has a model-following two-degree-of-freedom control configuration that is configured to follow the response y of the normative model-type real FB control system 7 having a transmission characteristic of rs → y. Therefore, the FB controller Cb plays the role of an FF controller by setting the model response ym to an ideal gain setting that matches the actual response y with a high response, and the FB controller CA plays the role of the model response ym and the actual response y. It plays a role of compensating for the deviation from. The FB controller CA can be designed so that the deviation y-ym can be suppressed stably and robustly independently of the FB controller Cb from the viewpoint of two-degree-of-freedom control.

As a result, the controller configuration of FIG. 1 can realize the model response ym-actual response y → 0 with high response, stability, and robustness.

Further, in FIG. 1, when CA = 0, it is equivalent to the generally known Smith method (when N = 1 in FIG. 6).

The yi shown in FIG. 1 is the predicted FB signal described in FIG. 3, and its role does not change even in the Smith method in which CA = 0. In the Smith method, the actual response y including the delay is offset by the model response ym, and the FB controller Cb is driven based on the predicted FB signal yi not including the delay to compensate for the delay. At this time, when the controlled object has a pole at the origin or the controlled object is unstable, it is difficult to cancel the actual response y with the model response ym. Therefore, in the Smith method, the controlled object is generally the origin. It was considered difficult to apply when there was a pole in the body or when the control target was unstable.

Also in the controller configuration of this embodiment, ym-y is formed by the adder / subtractor 5 and the adder / subtractor 6, and the FB controller Cb performs FB control based on the command value r-predicted FB signal yi. The idea of is the same as the Smith method. However, as described above, the controller configuration of this embodiment can realize the model response ym-actual response y → 0 with high response, stability, and robustness. Therefore, it can be considered that the controller configuration of this embodiment improves the canceling property of the actual response y by the model response ym, which is a drawback of the Smith method, by adopting the configuration of the model-following two-degree-of-freedom control.

In FIG. 2, which is a control system equivalent to FIG. 3, an arbitrary filter A (s) processes the FB signal of the normative model type real FB control system 27, and an arbitrary filter A (s) processes the model response. , The processing result is passed to the adder / subtractor 6. That is, the configuration is such that the model response ym-actual response y is processed by an arbitrary filter A (s), and the canceling characteristic that cancels the actual response y by the model response ym can be adjusted by an arbitrary filter A (s). be. However, when A (s) ≠ 1, it is necessary to separately filter the model response ym and the actual response y by A (s) in the configuration of FIG. 2, which is not advantageous in terms of calculation cost. In this case, the configuration shown in FIG. 3, which has the same transmission characteristics as that of FIG. 2, is adopted. In the configuration of FIG. 3, the filter A (s) is calculated only once for the model response ym − actual response y, and it is possible to avoid duplication of calculation costs for A (s). Therefore, when A (s) ≠ 1, adopting the configuration of FIG. 3 is more advantageous than the configuration of FIG. 2 in terms of calculation cost.

From the above, regarding the filter N designed in Patent Document 1, for example, when N represented by the equations (1) and (2) is adopted, the filter N can be implemented by mounting with the configuration of FIG. 1 or FIG. It is possible to reduce the calculation cost as compared with the implementation with the configuration shown in FIG. 6 described in Document 1. Further, since FIG. 1 or FIG. 3 has a model-following type 2 degree of freedom control configuration and the model response ym-actual response y → 0 can be realized with high response, stability, and robustness, the configuration of FIG. 1 or FIG. 3 is used. For example, it is possible to improve the canceling property of the actual response y by the model response ym, which was a drawback of the Smith method, and it is possible to compensate for the delay with high response and stability even when the controlled object has a pole at the origin. Is.

Next, how much the delay compensator 2 of the present embodiment shown in FIG. 3 can reduce the calculation cost as compared with the conventional delay compensator 62 will be described with a simple example.

In the FB control system shown in FIG. 6, it is assumed that the filter 61 adopted by the equation (1) is adopted. The FB control system equivalent to this is the FB control system including the delay compensator 42 shown in FIG. 4 as described above. The FB controller Cb is a PI controller, and the FB controller Cb can adjust the control response with the response frequency ωb as shown in the following equation (12).

However, N is an arbitrary positive real number.

Further, for the sake of simplicity, consider a case where the transmission characteristics of the nominal delay element model 15 are approximated by the Padé approximation (first order) as shown in the following equation (13).

Further, it is assumed that any FB control Ca included in the equation (1) has the same structure as the FB controller Cb. That is, the assumption of Eq. (3) is provided, and the FB controller Ca can adjust the controllability independently of ωb at the response frequency ωa.

Under the above assumption, the filter of the equation (1) becomes the following equations (14), (15), and (16).

In FIG. 6, when the model error signal ye is obtained from the operation amount u of the FB controller 16 and the output of the controlled object, the output yb of the delay compensator 62 is calculated by the following equation (17).

On the other hand, in FIG. 4, based on the premise that the output of the FB controller 3 has already been calculated as the previous state quantity, the model error signal ye is obtained from the manipulated variable u of the FB controller 16 and the output of the controlled object. If so, the output yb of the delay compensator 2 is calculated by the following equation (18).

Comparing Eqs. (17) and (18), the second term on the right side is different, and the second term on the right side of Eq. (18) has a lower degree of transfer function. Therefore, it can be seen that the delay compensator 42 of this embodiment has a lower calculation cost for calculating the output signal yb than the conventional delay compensator 62.

In this example, the Padé approximation is set to the first order, but in the case of a higher-order Pade approximation, the order of the transfer function of the second term on the right side of Eq. (17) increases, so the delay compensator 42 of this embodiment is used. It can be seen that the reduction of the calculation cost becomes more effective.

A filter N other than the equations (1) and (2) designed in Patent Document 1 is shown in FIG. 3 when the filter N is designed to be expressed as the following equation (19). With the configuration of the delay compensator 2, compensation characteristics equivalent to those of the delay compensator 62 shown in FIG. 6 can be realized, and the calculation cost can be reduced as compared with the delay compensator 62 shown in FIG.

For example, the filter A (s) of the delay compensator 2 in FIG. 3 is set as shown in the equation (20).

When defined by the constant γ, the delay compensator 52 shown in FIG. 5 is obtained. Although the proof is omitted, even with this configuration, it is possible to suppress the step disturbance applied to the input end of the controlled object without steady deviation when the controlled object has a pole at the origin.

As described above, according to the present embodiment, the delay compensator 2 shown in FIG. 3 can exhibit compensation performance equivalent to that of the delay compensator 62 adopting various filters N designed by the prior art. It can be seen that the configuration has the property and can reduce the calculation cost as compared with the delay compensator 62 of the prior art.

That is, when the delay compensator 2 shown in FIG. 3 is mounted on hardware such as a digital arithmetic unit, the calculation cost of the hardware applied to the filter 61 of the delay compensator 62 is not deteriorated as compared with the conventional technique. It is a control method suitable for mounting that can reduce the number of problems.

Further, according to this control method, it becomes possible to provide a control device using hardware that is cheaper and has lower calculation performance.

If the hardware such as the digital arithmetic unit to be mounted has sufficient arithmetic resources, it is described in Patent Document 1 as a delay compensator showing compensation performance equivalent to that of the delay compensator 2 shown in this embodiment. The FB control system may be implemented as shown in FIG. 6 by using the filter 61 designed by the technique.

As described above, according to the present embodiment, the processing of the filter 61 of the delay compensator 62 designed in Patent Document 1 is the exp (−τm · s) included in the denominator of the filter 61 and the molecule of the filter 61. -Since the calculation of the nominal plant model Pm included in the denominator can be realized by a simple control block configuration common to the calculation processing of the nominal delay element model 15 and the nominal plant model 14 required by the Smith method. It is possible to provide a feedback control method suitable for mounting, which can reduce the calculation cost in hardware mounting while maintaining the features and advantages of the delay compensator 62 designed in Patent Document 1, and a motor control device provided with the feedback control method. ..

The motor control method and the motor control device according to this embodiment assume the speed control system 71 in the cascade FB control system of the AC servomotor shown in FIG. 7, and the model to be controlled is specifically described in the following equation (21). It shall be as shown in (22) and (23).

Psm is a model that idealizes a nominal plant model in a speed control system, Mi is a model that idealizes a current control system that is a minor loop control system in a speed control system, and τsm is included in a closed loop of a current control system and a speed control system. The sum of all delays. Further, J, Ka, and Pp are inertia, motor constant, and pole logarithm, respectively, and ωi is the response frequency of the current control system.

Equations (21) to (23) indicate that it is necessary to deal with the problem that the controlled object has a pole at the origin in the design of the speed control system in the cascade FB control system of the AC servomotor.

The speed FB controller 72 of the speed control system is a PI controller, and has the following equations (24), (25), and (26).

However, L and ωs are the break point ratio and the response frequency of the speed control system, respectively. Generally, in order to regard the current control system as approximately 1, ωi is set to about several to 10 times ωs.

If ωs is increased to increase the response of the speed control system, the current control system cannot be approximated to 1 unless ωi is increased at the same time, and it must be regarded as a delay factor. In this case, the current control system is a first-order lag element as shown in the equation (23), and it is necessary to consider this as a lag element.

In this embodiment, the current control system is regarded as a delay element. In this problem setting, in Patent Document 1, the delay compensator 82 shown in FIG. 8 is configured, and the following equation (27) is given as a design example of the filter 81.

However, Csa is, as shown in the formula (28), as shown in the formula (28).

Although it has the same structure as the speed FB controller 86, its control design parameters are determined independently of the speed FB controller 86.

According to the prior art, the denominator of the equation (27) includes the response Mi of the current control system in addition to exp (−τm · s), and the order of the transfer function of the filter of the equation (27) tends to be high. be. Therefore, when the equation (27) is processed by the digital arithmetic unit 713, a high arithmetic cost is required.

FIG. 9 shows the configuration of the speed control system in the cascade FB control system of the AC servomotor in this embodiment. In FIG. 9, the delay compensator 92 in the present embodiment has the operation amount u output by the speed FB controller 86 and the rotation speed of the AC servomotor which is the control target with respect to the operation amount u output by the speed FB controller 86. It has a sensor detection value y as an input and outputs a speed FB signal yb with respect to the target rotation speed r.

The operation of the delay compensator 92 will be described with reference to FIG. The delay compensator 92 rotates the model to be controlled with respect to the signal obtained by adding the output of the arbitrary speed FB controller 93 to the AC servomotor to be controlled and the operation amount output by the speed FB controller 86 by the addition / subtractor 94. The model error signal ye is calculated by subtracting the speed response from the sensor detection value y of the rotational speed of the AC servomotor, which is the control target for the manipulated variable output by the FB controller 86, by the adder / subtractor 89.

Further, the delay compensator 92 calculates the response of the plant nominal model 84 to the signal obtained by adding the output of the arbitrary speed FB controller 93 and the operation amount output by the speed FB controller 86 by the addition / subtractor 94 as the predicted FB signal. Then, the signal obtained by processing the model error signal by the filter 91 and the predicted FB signal are added by the adder / subtractor 95, and output as a speed FB signal yb with respect to the target rotation speed r.

The speed FB controller 86 receives a deviation signal obtained by subtracting the speed FB signal yb output by the delay compensator 92 from the target rotation speed r by the adder / subtractor 87 as an input, and generates an operation amount u for the controlled object.

In the delay compensator 92 shown in FIG. 9 in this embodiment, CsA and As (s) are designed as, for example, the following equations (29) and (30).

However, Csa in the formula (29) is represented by the formula (28).

The delay compensator 92 designed in this way has input / output characteristics equivalent to those of the delay compensator 82 including the filters N represented by the equations (27) and (28) designed in Patent Document 1 from a simple calculation. Can be confirmed to indicate. That is, the delay compensator 92 designed as shown in the equations (29) and (30) is equivalent to the delay compensator 82 including the filters of the equations (27) and (28) designed in Patent Document 1. Has performance.

Both equations (29) and (30) include the transfer characteristic Pm of the nominal plant model 84 included in the equation (27) and the transfer characteristic Mi · exp (−τm · s) of the nominal delay element model 85. It can be seen that it is a simple transfer function.

In this way, the reason why the transfer functions of CsA and As (s) are simplified as compared with the equation (27) is that the Pm and Mi · exp (−τm · s) contained in the equation (27). This is because the delay compensator 92 is configured as a block configuration so that the calculation can be shared with the calculation of the nominal plant model 84 and the nominal delay element model 85.

As described above, according to the present embodiment, the delay compensator 92 in the speed control system of the AC servomotor shown in FIG. 9 is equivalent to the delay compensator 82 using various filters N designed by the prior art. The configuration is such that the compensation performance can be shown and the calculation cost can be reduced as compared with the delay compensator 82 of the prior art.

That is, when the delay compensator 92 shown in FIG. 9 is mounted on the digital arithmetic unit 713, the arithmetic cost of the digital arithmetic unit 713 applied to the filter 81 of the delay compensator 82 is reduced without deterioration of the control performance as compared with the prior art. It is a control method suitable for mounting that can be reduced.

Further, according to this control method, it becomes possible to provide a control device by a digital arithmetic unit 713 which is cheaper and has low arithmetic performance.

In the speed control system of the AC servomotor, the filter 81 of the delay compensator 82 is described by the following equation (31), using the prior art.

Even when the design is as follows, in the present embodiment, in the delay compensator 92 shown in FIG. 9, the following equation (32),

If the design is as follows, the delay compensator 92 shown in FIG. 9 and the delay compensator 82 including the filter of the equation (31) have equivalent compensation characteristics.

In this, for example, as in the following equation (33),

Even when defined by the constant γ, when the controlled object has a pole at the origin, it is possible to suppress the step disturbance applied to the input end of the controlled object without steady deviation. If it is the equation (33), it can be said that the delay compensator 92 adopting the equation (28) is a suitable control method for mounting without any particular increase in calculation cost from the comparison with the equation (30).

Further, according to this control method, it becomes possible to provide a control device by a digital arithmetic unit 713 which is cheaper and has low arithmetic performance.

1,61,81,91: Filter 2,42,52,62,82,92: Delay compensator, 3,16,93: Feedback controller, 4,5,6,18,35,37,39: Add / subtractor, 7, 27: Normal model type real feedback control system, 8: Model feedback control system, 12: Control target, 13: Feedback delay element, 14: Nominal plant model, 15: Nominal delay element model, 71: AC Servo motor speed control system, 77: AC servo motor, 78: Current detector, 713: Digital arithmetic unit

Claims (6)

A feedback control method for a feedback control system including a first delay compensator composed of a model to be controlled, an arbitrary first filter, and an arbitrary first feedback controller.
The controlled model includes a nominal plant model that simulates the dynamics of the controlled object and a nominal delay model that simulates a delay element contained in the closed loop of the feedback control system.
The first delay compensator uses an operation amount output by a second feedback controller that performs feedback control on the controlled object and an output signal of the controlled object as input signals.
The output signal of the nominal plant model with respect to the signal obtained by adding the operation amount and the output of the arbitrary first feedback controller included in the first delay compensator by an adder / subtractor is used as an ideal feedback signal, and the operation amount is used. And the output signal of the model to be controlled with respect to the signal obtained by adding / subtracting the output of the arbitrary first feedback controller included in the first delay compensator, and the output signal of the controlled object are added / subtracted. The signal obtained by subtraction by the device is used as a model error signal, the model error signal is used as an input of the arbitrary first feedback controller, and the model error signal is used as a signal processed by the arbitrary first filter. The signal obtained by adding the ideal feedback signal with the adder / subtractor is used as the output signal.
The second feedback controller calculates the deviation between the output signal of the first delay compensator and the target value signal by an adder / subtractor, and performs feedback compensation to the controlled object based on the deviation. A feedback control method characterized by. The feedback control method according to claim 1.
The feedback control method, wherein the arbitrary first filter is an arbitrary real number whose transfer characteristic is larger than 0. The feedback control method according to claim 1.
The feedback control method, wherein the optional first filter has a transmission characteristic of 1. A motor control device that employs a feedback control system that includes a first delay compensator composed of a model to be controlled, an arbitrary first filter, and an arbitrary first feedback controller.
The controlled model includes a nominal plant model that simulates the dynamics of the controlled object and a nominal delay model that simulates a delay element contained in the closed loop of the feedback control system.
The first delay compensator uses an operation amount output by a second feedback controller that constitutes the feedback control system and performs feedback control on a controlled object and an output signal of the controlled object as input signals.
The output signal of the nominal plant model with respect to the signal obtained by adding the operation amount and the output of the arbitrary first feedback controller included in the first delay compensator by an adder / subtractor is used as an ideal feedback signal, and the operation amount is used. And the output signal of the model to be controlled with respect to the signal obtained by adding / subtracting the output of the arbitrary first feedback controller included in the first delay compensator, and the output signal of the controlled object are added / subtracted. The signal obtained by subtraction by the device is used as a model error signal, the model error signal is used as an input of the arbitrary first feedback controller, and the model error signal is used as a signal processed by the arbitrary first filter. The signal obtained by adding the ideal feedback signal with the adder / subtractor is used as the output signal.
The second feedback controller calculates the deviation between the output signal of the first delay compensator and the target value signal by an adder / subtractor, and performs feedback compensation to the controlled object based on the deviation. A motor control device characterized by. The motor control device according to claim 4.
The optional first filter is a motor control device whose transmission characteristic is an arbitrary real number larger than 0. The motor control device according to claim 4.
The optional first filter is a motor control device having a transmission characteristic of 1.

Images