JP7769100B2 - Adjustment device for adjusting control parameters, control system, and control parameter adjustment method - Google Patents

Description

The present invention relates to an adjustment device that adjusts control parameters of a motor control device that controls a motor, a control system including the adjustment device, and a control parameter adjustment method.

When optimizing control parameters such as motor gain and filter coefficients to satisfy pre-set stability conditions (such as stability margins) for motors used to drive machines such as machine tools, robots, or industrial machinery, it was necessary to perform a series of adjustment processes multiple times, including setting the control parameters, operating the motor control device to measure the machine's frequency characteristics, adjusting the control parameters, operating the motor control device to measure the frequency characteristics again, and confirming whether the stability conditions were satisfied.

Patent Document 1 describes a control parameter sensitivity analysis device for checking the results of control parameter adjustment in order to optimally adjust the control parameters of a motor control device used in semiconductor manufacturing equipment, machine tools, or industrial robots.
Patent Document 1 describes that a control parameter sensitivity analysis device is provided with an open-loop frequency response characteristic measurement means for acquiring an open-loop frequency response characteristic that does not include the characteristics of the controller, a control model of the controller of the motor control device, a calculation means for calculating a closed-loop frequency response characteristic from the measured open-loop frequency response characteristic and the control model, and a sensitivity analysis device for performing sensitivity analysis on the relationship between the control parameters of the controller and changes in the closed-loop frequency response characteristic, in an electric motor control device comprising a detection means for detecting an operating amount of a machine, a command device for generating a command signal, and a controller for receiving the command signal and driving the electric motor.

Patent document 2 describes a control assistance device that adjusts at least one of the coefficient of at least one filter and the feedback gain (FG) (which become control parameters) of a servo control unit (which becomes a motor control unit) that controls a motor.
Patent document 2 describes that a control assistance device uses second information and first information consisting of at least one filter coefficient and feedback gain of a servo control device that controls a motor before and after adjustment to calculate at least one frequency characteristic of the input/output gain and phase delay between the filter and the feedback gain before and after adjustment of at least one of the filter coefficient and feedback gain, and obtains an estimated value of the frequency characteristic of the input/output gain and phase delay of the servo control device after adjustment of at least one of the filter coefficient and feedback gain based on at least one frequency characteristic before and after adjustment and the measured frequency characteristics of the input/output gain and input/output phase delay of the servo control device before adjustment of at least one of the coefficient and feedback gain.

Japanese Patent Application Laid-Open No. 2005-275588 International Publication No. 2021/251226

When setting a stability margin as a stability condition for a motor control device, if the stability margin (which refers to a gain margin and a phase margin) is large, stability increases but responsiveness decreases. On the other hand, if the stability margin is small, stability decreases but responsiveness increases.
When setting the stability margin of a motor control device, a user may want to determine the adjustment conditions based on how the frequency characteristics of the motor control device change under multiple adjustment conditions, such as standard, stability-oriented, responsiveness-oriented, and custom.
However, it takes a lot of time to repeat the process of adjusting control parameters such as motor gain and filter coefficients to find frequency characteristics for each of a plurality of adjustment conditions.
Therefore, it is desirable to be able to confirm the results of adjusting the control parameters of a motor control device under multiple adjustment conditions by measuring the frequency characteristics once.

(1) A first aspect of the present disclosure is an adjustment device that adjusts control parameters of a motor control unit that controls a motor, comprising:
a frequency characteristic storage unit that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit that stores the plurality of control parameters optimized for the plurality of adjustment conditions;
an evaluation index calculation unit that calculates an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameter;
a presentation unit that presents at least one of the predicted frequency characteristics and the evaluation index corresponding to the optimized control parameters for each of a plurality of adjustment conditions;
a control parameter setting unit that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
It is an adjustment device equipped with:

(2) A second aspect of the present disclosure is an adjustment device that adjusts control parameters of a motor control unit that controls a motor, comprising:
a frequency characteristic storage unit that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit that stores the plurality of control parameters optimized for the plurality of adjustment conditions;
a time response prediction unit that predicts a first time response using a predicted frequency characteristic corresponding to the optimized control parameter;
an evaluation index calculation unit that calculates an evaluation index of the first time response from the predicted first time response;
a presentation unit that presents at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
a control parameter setting unit that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
It is an adjustment device equipped with:

(3) A third aspect of the present disclosure is a control system comprising a motor control unit that controls a motor and an adjustment device described above in (1) or (2).

(4) A fourth aspect of the present disclosure is a control parameter adjustment method for adjusting control parameters of a motor control unit that controls a motor, the method comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
A process of calculating an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameters;
a process of presenting at least one of the predicted frequency characteristics and the evaluation index corresponding to the optimized control parameters for each of the plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
This is a control parameter adjustment method that performs the above.

(5) A fifth aspect of the present disclosure is a control parameter adjustment method for adjusting control parameters of a motor control unit that controls a motor, the method comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
predicting a first time response using the predicted frequency characteristics corresponding to the optimized control parameters;
calculating an evaluation index of the first time response from the predicted first time response;
a process of presenting at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
This is a control parameter adjustment method that performs the above.

According to each aspect of the present disclosure, a single measurement of frequency characteristics can determine multiple frequency characteristics when control parameters such as the gain of the motor control unit and filter coefficients are adjusted under multiple adjustment conditions. As a result, by checking multiple frequency characteristics and/or evaluation indices of multiple frequency characteristics, it is possible to easily compare the frequency characteristics and/or evaluation indices of frequency characteristics after adjustment under different adjustment conditions and easily select the control parameters to be applied. Furthermore, by checking multiple time responses and/or evaluation indices of multiple time responses predicted from multiple frequency characteristics, it is possible to easily compare the time responses and/or evaluation indices of time responses after adjustment under different adjustment conditions and easily select the control parameters to be applied.

FIG. 1 is a block diagram illustrating a control system according to a first embodiment of the present disclosure. 1 is a Bode diagram showing a gain margin, a phase margin, a maximum gain of a closed loop characteristic, and a maximum gain in a high frequency region. FIG. 2 is a block diagram showing an example of the configuration of an adjustment unit. FIG. 10 is a diagram showing an example of a setting screen for adjustment conditions. 1 is a diagram showing a complex plane showing a unit circle on the complex plane and two circles that form closed curves. FIG. 1 is a diagram illustrating a Nyquist locus, a unit circle, and a circle passing through a gain margin and a phase margin, drawn on a complex plane. FIG. 10 is a diagram showing a display screen displaying Bode diagrams of open-loop frequency characteristics or closed-loop frequency characteristics for “before adjustment,” “standard,” and “stability-oriented,” and evaluation indexes for “before adjustment,” “standard,” and “stability-oriented.” A Bode diagram shown in display field 502A. This is a Bode diagram shown in display field 502B. This is a Bode diagram shown in display field 502C. 10A and 10B are Bode diagrams showing open-loop or closed-loop frequency characteristics for "standard," "stability-oriented," "response-oriented," and "custom." FIG. 10 is a diagram showing a setting screen for adjustment conditions. 10 is a flowchart showing the operation of an adjustment unit. FIG. 10 is a block diagram illustrating a modified example of the adjustment unit in which the control parameter adjustment unit is replaced with a machine learning unit. FIG. 2 is a block diagram showing the configuration of a machine learning unit. FIG. 1 is a block diagram showing a reference model of a closed loop. FIG. 10 is a characteristic diagram showing frequency characteristics of input/output gains of a motor control unit of a reference model and the motor control units before and after learning. FIG. 10 is a block diagram showing a configuration example of an adjustment unit included in a control system according to a second embodiment of the present disclosure. FIG. 4 is a Bode diagram showing a first resonance mode and a second resonance mode. FIG. 10 is a diagram showing a display screen displaying time responses for "before adjustment," "standard," and "stability-oriented" and evaluation indices for "before adjustment," "standard," and "stability-oriented." FIG. 7 is a diagram showing the time response characteristics shown in display field 702A. FIG. 7 is a diagram showing the time response characteristics shown in display field 702B. FIG. 7 is a diagram showing the time response characteristics shown in display field 702C. FIG. 10 is a block diagram showing an example of a filter configured by directly connecting a plurality of filters. FIG. 10 is a block diagram showing another example of the configuration of the control system.

Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings.
(First embodiment)
FIG. 1 is a block diagram showing a control system according to a first embodiment of the present disclosure.
The control system 10 includes a motor control unit 100, a frequency generation unit 200, a frequency characteristic measurement unit 300, and an adjustment unit 400. The motor control unit 100 corresponds to a motor control device, and the adjustment unit 400 corresponds to an adjustment device.
One or more of the frequency generating section 200, the frequency characteristic measuring section 300, and the adjusting section 400 may be provided within the motor control section 100. The frequency characteristic measuring section 300 may be provided within the adjusting section 400.

The motor control unit 100 includes a subtractor 110, a speed control unit 120, a filter 130, a current control unit 140, and a motor 150. The subtractor 110, the speed control unit 120, the filter 130, the current control unit 140, and the motor 150 form a servo system with a closed speed feedback loop. The motor 150 may be a linear motor that moves in a straight line, or a motor with a rotating shaft. The object driven by the motor 150 is, for example, a mechanical part of a machine such as a machine tool, a robot, or industrial machinery. The motor 150 may be provided as part of a machine tool, a robot, industrial machinery, etc. The control system 10 may be provided as part of a machine tool, a robot, industrial machinery, etc. Details of the configuration of the motor control unit 100 will be described later.

The frequency generating unit 200 outputs a sine wave signal as a speed command while changing the frequency to the subtractor 110 and frequency characteristic measuring unit 300 of the motor control unit 100.

The frequency characteristic measuring unit 300 uses the speed command (sine wave) generated by the frequency generating unit 200 as an input signal, and the detected speed (sine wave) as an output signal output from a rotary encoder (not shown) provided on the motor 150 or the integral of the detected position (sine wave) as an output signal output from the linear scale, to determine the frequency characteristics of the amplitude ratio (input/output gain) between the input signal and the output signal and the phase delay for each frequency defined by the speed command, and outputs the determined frequency characteristics to the adjusting unit 400. The determined frequency characteristics are the closed-loop frequency characteristics Pc.
Furthermore, the frequency characteristic measuring unit 300 calculates the open-loop frequency characteristic Po from the frequency characteristic Pc and outputs it to the adjustment unit 400. The closed-loop frequency characteristic Pc is expressed as Pc = Po/(1 + Po) using the open-loop frequency characteristic Po. Therefore, the open-loop frequency characteristic Po can be obtained as Po = Pc/(1 - Pc).

For each of the plurality of adjustment conditions, the adjustment unit 400 determines an optimal value for one or both of the integral gain K1v and proportional gain K2v of the speed control unit 120, and at least one of the coefficients _ωc , τ, and δ (which serve as control parameters) of the transfer function of the filter 130. The plurality of adjustment conditions are, for example, two or more of adjustment conditions classified according to frequency response characteristics, such as standard, stability-oriented, responsiveness-oriented, and custom. "Standard" is intermediate between stability-oriented and responsiveness-oriented, and "Custom" is set arbitrarily by the user.
Each adjustment condition is a condition that imposes a limit on at least one of the gain margin, the phase margin, the maximum gain of the closed-loop characteristics, and the maximum gain in the high-frequency region.
FIG. 2 is a Bode diagram showing the gain margin, the phase margin, the maximum gain of the closed loop characteristics, and the maximum gain in the high frequency region.
Table 1 shows examples of limit values for the gain margin, phase margin, maximum gain of the closed loop characteristics, and maximum gain in the high frequency region for standard, stability-oriented, response-oriented, and custom.

The adjustment unit 400 displays the frequency characteristics of the machine for the optimum control parameters found for each adjustment condition, or calculates and displays evaluation indexes such as the gain margin, phase margin, and control band of the frequency characteristics. The adjustment unit 400 then allows the user to select a frequency characteristic or an evaluation index from the multiple frequency characteristics or multiple evaluation indexes displayed for each adjustment condition, and sets the optimum control parameters corresponding to the selected frequency characteristic or evaluation index, i.e., one or both of the integral gain K1v and proportional gain K2v of the speed control unit 120, and at least one of the coefficients _ωc , τ, and δ of the transfer function of the filter 130, in the motor control unit 100.

The motor control unit 100 and the adjustment unit 400 will be described in more detail below.
<Motor control unit 100>
As already described, the motor control unit 100 includes the subtractor 110 , the speed control unit 120 , the filter 130 , the current control unit 140 , and the motor 150 .

The subtractor 110 calculates the difference between the input speed command and the detected speed fed back, and outputs this difference to the speed control unit 120 as the speed deviation.

The speed control unit 120 performs PI control (Proportional-Integral Control), adds together a value obtained by multiplying the speed deviation by an integral gain K1v and integrating the result, and a value obtained by multiplying the speed deviation by a proportional gain K2v, and outputs the result as a torque command to the filter 130. The speed control unit 120 includes a feedback gain. Note that the speed control unit 120 is not limited to PI control, and other control, such as PID control (Proportional-Integral-Differential Control), may also be used.
Equation 1 (hereinafter referred to as Equation 1) represents the transfer function G _V (s) of the speed control unit 120 .

The filter 130 is a filter that attenuates specific frequency components, and may be, for example, a notch filter, a low-pass filter, or a band-stop filter. A resonance point exists in a machine tool or other such machine that has a mechanical part driven by the motor 150, and resonance may increase in the motor control unit 100. Resonance can be reduced by using a filter such as a notch filter. The output of the filter 130 is output to the current control unit 140 as a torque command.
Equation 2 (hereinafter referred to as Equation 2) represents the transfer function G _F (s) of the notch filter serving as filter 130. The parameters represent coefficients ω _c , τ, and δ.
In Equation 2, the coefficient δ is the attenuation coefficient, the coefficient _ωc is the central angular frequency, and the coefficient τ is the fractional bandwidth. If the central frequency is fc and the bandwidth is fw, the coefficient _ωc is expressed as _ωc =2πfc, and the coefficient τ is expressed as τ=fw/fc.

The current control unit 140 generates a current command for driving the motor 150 based on the torque command, and outputs the current command to the motor 150 .
When motor 150 is a linear motor, the position of the movable part is detected by a linear scale (not shown) provided on motor 150, and a speed detection value is obtained by differentiating the position detection value. The obtained speed detection value is input to subtractor 110 as speed feedback.
If the motor 150 is a motor having a rotating shaft, the rotation angle position is detected by a rotary encoder (not shown) provided on the motor 150, and the detected speed value is input to the subtractor 110 as speed feedback.

<Adjustment section 400>
3 is a block diagram showing an example of the configuration of the adjustment unit 400. As shown in FIG. 3, the adjustment unit 400 includes a frequency characteristic storage unit 401, an adjustment condition setting unit 402, a frequency characteristic prediction unit 403, a control parameter adjustment unit 404, a control parameter storage unit 405, an evaluation index calculation unit 406, a presentation unit 407, and a control parameter setting unit 408.
Each part of the adjustment unit 400 will be described below.

(Frequency characteristic storage unit 401)
The frequency characteristic storage unit 401 stores the closed-loop frequency characteristic Pc and the open-loop frequency characteristic Po of the input/output gain and phase delay output from the frequency characteristic measurement unit 300. The closed-loop frequency characteristic Pc is the frequency characteristic of the machine acquired by the frequency characteristic measurement unit 300 when the motor control unit 100 operates using the control parameters before adjustment.

(Adjustment condition setting unit 402)
The adjustment condition setting unit 402 displays a setting screen for inputting adjustment conditions, and sets at least one (stability condition) of the gain margin, phase margin, maximum gain of the closed-loop characteristics, and maximum gain of the high-frequency region of the open-loop circuit of the motor control unit 100 for each of the multiple adjustment conditions input by the user. The gain margin, phase margin, maximum gain of the closed-loop characteristics, and maximum gain of the high-frequency region may be set to predetermined values in advance, or may be set arbitrarily by the user. The following explanation will be given for the case where the stability margins (gain margin and phase margin) are set.
When a plurality of adjustment conditions are input by the user, the adjustment condition setting unit 402 notifies the control parameter adjustment unit 404 that a plurality of adjustment conditions have been input.

The adjustment condition setting unit 402 outputs image data including a closed curve passing through the gain margin and the phase margin on the complex plane to the control parameter adjustment unit 404 for each adjustment condition, based on a request from the control parameter adjustment unit 404. The open-loop circuit is made up of the speed control unit 120, the filter 130, the current control unit 140, and the motor 150 shown in FIG.
Specifically, the adjustment condition setting unit 402 first displays the setting screen shown in Fig. 4, and the user selects adjustment conditions on the setting screen. The user selects two or more adjustment conditions for the cutting feed and rapid traverse of the X-axis and Y-axis, respectively, from the four adjustment conditions shown in Fig. 4, for example, standard, stability emphasis, responsiveness emphasis, and custom. Fig. 4 shows how the user sequentially selects and sets standard and stability emphasis for rapid traverse of the Y-axis.

Next, the adjustment condition setting unit 402 draws a unit circle on the complex plane in Figure 5 whose circumference passes through (-1,0), and based on the stability margin based on the adjustment conditions, draws a closed curve such as a circle that intersects with this unit circle and includes (-1,0) on the complex plane within it, and outputs image data including the closed curve to the control parameter adjustment unit 404 for each adjustment condition. It is not necessary for image data to be output to the control parameter adjustment unit 404; it is sufficient that the data output to the control parameter adjustment unit 404 at least represents a closed curve such as a circle on the complex plane. In the following explanation, it is assumed that the adjustment condition setting unit 402 outputs image data to the control parameter adjustment unit 404.

FIG. 5 shows the complex plane when the adjustment condition is "standard" and when "stability is emphasized."
When the adjustment condition is "standard," the adjustment condition setting unit 402 draws the small-diameter circle C2 as a closed curve on the complex plane and outputs the image data to the control parameter adjustment unit 404. When the adjustment condition is "stability-oriented," the adjustment condition setting unit 402 draws the large-diameter circle C1 as a closed curve on the complex plane and outputs the image data to the control parameter adjustment unit 404.
The point where circle C1 or C2 intersects with the real axis determines the gain margin, and the point where circle C1 or C2 intersects with the unit circle determines the phase margin. As the diameter of the circle increases, the stability margin (gain margin and phase margin) also increases, increasing stability but decreasing responsiveness.

In Figure 5, the centers of circles C1 and C2 are on the real axis, but they do not have to be on the real axis. The closed curve may be a closed curve other than a circle, such as a rhombus, rectangle, or ellipse. In Figure 5, circles C1 and C2 have the same center, but the center of circle C1 and the center of circle C2 may be different.

(Frequency characteristic prediction unit 403)
The frequency characteristic prediction unit 403 stores the transfer function G _V (jω) of the speed control unit 120 and/or the transfer function G _F (jω) of the filter 130 that use the control parameters before adjustment.
The unadjusted control parameters are generated in advance by the user. If the operator has adjusted the control parameters in advance, the adjusted values may be used as the "unadjusted control parameters."
Then, the frequency characteristic prediction unit 403 calculates frequency characteristics C1 of the input/output gain and phase delay of the speed control unit 120 and/or the filter 130 using the transfer function G _V (jω) of the speed control unit 120 and/or the transfer function G _F (jω) of the filter 130, which are obtained using the control parameters before adjustment. Furthermore, the frequency characteristic prediction unit 403 calculates frequency characteristics _C2 of the input/output gain and phase delay of the speed control unit 120 and/or the filter 130, which are obtained using the transfer function G _V (jω) of the speed control unit 120 and/or the transfer function G _F (jω) of the filter 130, which are obtained using the adjusted control parameters that have been adjusted based on the adjustment information output from the control parameter adjustment unit _404. The transfer function G _V (jω) of the speed control unit 120 and/or the transfer function G _F (jω) of the filter 130, which are obtained using the adjusted control parameters, can be obtained by replacing the control parameters before adjustment with the adjusted control parameters.

For example, when the integral gain K1v and proportional gain K2v of the transfer function _GV (jω) of the speed control unit 120 are multiplied by n (n is an integer) from the pre-adjustment values, the frequency characteristics are n× _GV (jω), where _GV (jω) is the pre-adjustment frequency characteristic. On the Bode diagram, the gain and phase are respectively expressed by Equation 3 (hereinafter referred to as Equation 3) and Equation 4 (hereinafter referred to as Equation 4).

In this manner, the frequency characteristic prediction unit 403 obtains the calculated frequency characteristics _C1 and _C2 .

The frequency characteristic prediction unit 403 further performs the following processing using the calculated frequency characteristics _C1 and _C2 .
The frequency characteristic prediction unit 403 calculates an estimated value Eo of the open loop frequency characteristics of the input/output gain and phase delay of the motor control unit 100 based on the frequency characteristics C ₁ , the frequency characteristics C ₂ and the open loop frequency characteristics Po acquired from the frequency characteristic storage unit 401 .
Specifically, the frequency characteristic prediction unit 403 uses the following equation 5 (hereinafter referred to as equation 5) to obtain an estimated value Eo of the open-loop frequency characteristics of the input/output gain and phase delay of the motor control unit 100.

Furthermore, the frequency characteristic prediction unit 403 calculates the estimated value Ec of the closed loop frequency characteristic by using the estimated value Eo of the open loop frequency characteristic, Ec=Eo/(1+Eo).
The estimated value Eo of the frequency characteristics of the input/output gain and phase delay of the motor control unit 100 can be calculated using the above formula 5, i.e., Eo = C ₂ - _{C 1} + Po, but the calculation performed by the frequency characteristic prediction unit 403 may be any of Eo = (C ₂ - C ₁ ) + Po, Eo = (Po - C ₁ ) + C ₂ , or E = (Po + C ₂ ) - C ₁ .
The case where the frequency characteristic prediction unit 403 obtains the estimated value Eo of the open-loop frequency characteristic by the equation Eo=(C ₂ −C ₁ )+Po will be further described below.

For example, as described above, when the integral gain K1v and proportional gain K2v of the transfer function _GV (jω) of the speed control unit 120 are multiplied by n (n is an integer) from the values before adjustment, the frequency characteristic _C1 becomes n× _GV (jω), where _GV (jω) is the frequency characteristic _C2 in the initial state.
The frequency characteristic of the gain when multiplied by n from the initial state, as shown in the Bode plot, is the frequency characteristic obtained by adding 20log ₁₀ (n) to 20log ₁₀ |G _V (jω)|, which is the gain in the initial state, as shown in Equation 4. The frequency characteristic of the phase when multiplied by n from the initial state, as shown in the Bode plot, is tan ^-1 (n) = 0, and therefore is unchanged from the phase-frequency characteristic in the initial state, as shown in Equation 4.
Therefore, when the frequency characteristic is multiplied by n from the state before adjustment, only the gain diagram of the Bode plot changes, and 20log ₁₀ (n) shown in Equation 3 becomes the difference (C ₂ −C ₁ ) between the two frequency characteristics: the frequency characteristic when multiplied by n from the state before adjustment and the frequency characteristic in the state before adjustment.

The frequency characteristic prediction unit 403 reads the open-loop frequency characteristic Po of the input/output gain and phase delay from the frequency characteristic storage unit 401, and calculates an estimated value Eo of the open-loop frequency characteristic by adding the difference (C ₂ −C ₁ ) to this open-loop frequency characteristic Po. Then, the frequency characteristic prediction unit 403 calculates an estimated value Ec of the closed-loop frequency characteristic using the estimated value Eo of the open-loop frequency characteristic by Ec=((C ₂ −C ₁ )+Po)(1+(C ₂ −C ₁ )+Po).

The frequency characteristic prediction unit 403, in operation with the control parameter adjustment unit 404 described later, optimizes the control parameters for each adjustment condition, obtains the optimized open-loop frequency characteristic or closed-loop frequency characteristic (estimated value), and outputs it to the presentation unit 407 and the evaluation index calculation unit 406.
Specifically, the frequency characteristic prediction unit 403 outputs the open-loop frequency characteristic or the closed-loop frequency characteristic (estimated value) corresponding to the optimized control parameters for “standard” to the presentation unit 407 and the evaluation index calculation unit 406 .
In addition, the frequency characteristic prediction unit 403 outputs the open-loop frequency characteristic or closed-loop frequency characteristic (estimated value) output from the frequency characteristic prediction unit 403, which corresponds to the optimized control parameter for "stability emphasis", to the presentation unit 407 and the evaluation index calculation unit 406.
Furthermore, the frequency characteristic prediction unit 403 outputs the open-loop frequency characteristic or the closed-loop frequency characteristic output from the frequency characteristic storage unit 401 using the initial control parameters before adjustment to the presentation unit 407 and the evaluation index calculation unit 406 .

By using the frequency characteristic prediction unit 403 described above, it is possible to calculate an estimated value Ec of the closed-loop frequency characteristics of the input/output gain and phase delay of the motor control unit 100 using the adjusted control parameters, which can be obtained in a shorter time than when operating the motor control unit 100 using the adjusted control parameters, actually detecting the speed command and detected speed, and measuring the closed-loop frequency characteristics using the frequency characteristic measurement unit 300.

(Control parameter adjustment unit 404)
The control parameter adjustment unit 404 acquires image data including a closed curve passing through the gain margin and the phase margin on the complex plane from the adjustment condition setting unit 402 .
In the following explanation, it is assumed that the control parameter adjustment unit 404 first acquires image data in which a small-diameter circle C2 corresponding to "standard" is drawn as a closed curve on the complex plane, and then acquires image data in which a large-diameter circle C1 corresponding to "stability-oriented" is drawn as a closed curve on the complex plane.
In addition, the control parameter adjustment unit 404 obtains from the frequency characteristic prediction unit 403 the frequency characteristics (estimated values of open-loop frequency characteristics or estimated values of closed-loop frequency characteristics) of the input/output gain and phase delay of the motor control unit 100 based on the adjusted control parameters.

The control parameter adjustment unit 404 plots a Nyquist locus created from the open-loop frequency characteristic H(jω)' (estimated value) or the frequency characteristic G(jω)' (estimated value) of the closed-loop frequency characteristic output from the frequency characteristic prediction unit 403 on a complex plane including the unit circle and the circle C2, which is a closed curve. The method for creating the Nyquist locus will be described later.

FIG. 6 is a diagram showing the Nyquist locus, the unit circle, and the circle passing through the gain margin and the phase margin, drawn on the complex plane.
The control parameter adjustment unit 404 outputs adjustment information to the frequency characteristic prediction unit 403 to adjust at least one of the control parameters, the integral gain K1v and proportional gain K2v of the speed control unit 120, and the coefficients _ωc , τ, and δ of the transfer function of the filter 130, so that the Nyquist locus does not pass inside the circle C2.
The control parameters are optimized so that the Nyquist locus does not pass inside the circle C2 by repeatedly adjusting the control parameters between the frequency characteristic prediction unit 403 and the control parameter adjustment unit 404. The control parameter adjustment unit 404 stores the optimized control parameters in the control parameter storage unit 405 as control parameters related to "standard."

After determining the control parameters for "standard," the control parameter adjustment unit 404 acquires image data in which a large-diameter circle C1 corresponding to "stability emphasis" is drawn as a closed curve on a complex plane, and determines the control parameters for "stability emphasis" in the same way as the control parameters for "standard," and stores these in the control parameter storage unit 405.

A method for creating the Nyquist locus by the control parameter adjustment unit 404 will be specifically described below.
The frequency characteristic measuring unit 300 stores in the frequency characteristic storage unit 401 the closed-loop frequency characteristics measured by driving the motor control unit 100 using the initial control parameters before adjustment (integral gain K1v, proportional gain K2v, and coefficients _ωc , τ, δ) and the open-loop frequency characteristics calculated from these closed-loop frequency characteristics.
The frequency characteristic measuring section 300 calculates the open loop frequency characteristic from the closed loop frequency characteristic as follows.
The speed feedback loop is composed of a subtractor 110 and an open loop circuit with a transfer function H. As already explained, the open loop circuit is composed of the speed control unit 120, filter 130, current control unit 140, and motor 150 shown in FIG. 1. When the input/output gain of the speed feedback loop at a certain frequency _ω0 is c and the phase delay is θ, the closed loop frequency characteristic G( _jω0 ) is c·e ^jθ . Using the open loop frequency characteristic H( _jω0 ), the closed loop frequency characteristic G( _jω0 ) is expressed as G( _jω0 ) = H( _jω0 )/(1 + H( _jω0 )). Therefore, the open loop frequency characteristic H(jω ₀ ) at a certain frequency ω ₀ can be obtained by H(jω ₀ ) = G(jω ₀ )/(1 - G(jω ₀ )) = c·e ^jθ /(1 - c·e ^jθ ).

The frequency characteristic prediction unit 403 reads out the open-loop frequency characteristic Po (Po=H(jω)) or the closed-loop frequency characteristic Pc (Pc=G(jω)) based on the initial control parameters before adjustment from the frequency characteristic storage unit 401, and uses Equation 3 to determine the estimated value Eo of the open-loop frequency characteristic or the estimated value Ec of the closed-loop frequency characteristic when the adjusted control parameters are used, and outputs the estimated value Eo or the closed-loop frequency characteristic to the control parameter adjustment unit 404.
The control parameter adjustment unit 404 creates a Nyquist locus by plotting the estimated value Eo (Eo=H(jω)′) of the open-loop frequency characteristic or the estimated value Ec=G(jω)′ of the closed-loop frequency characteristic obtained from the frequency characteristic measurement unit 300 on a complex plane.

(Control parameter storage unit 405)
The control parameter storage unit 405 stores control parameters relating to "standard" and control parameters relating to "safety-oriented."

(Evaluation index calculation unit 406)
The evaluation index calculation unit 406 calculates an evaluation index related to the "before adjustment" and outputs it to the presentation unit 407. The evaluation index related to the "before adjustment" is, for example, at least one of three items: a gain margin and a phase margin calculated based on the open-loop frequency characteristics corresponding to the initial control parameters before adjustment, and a control band calculated based on the closed-loop frequency characteristics corresponding to the initial control parameters before adjustment.
The evaluation index calculation unit 406 calculates an evaluation index for the "standard" and outputs it to the presentation unit 407. The evaluation index for the "standard" is, for example, at least one of three items: a gain margin and a phase margin calculated based on the open-loop frequency characteristics corresponding to the control parameters for the "standard", and a control band calculated based on the closed-loop frequency characteristics corresponding to the control parameters for the "standard".
Furthermore, the evaluation index calculation unit 406 calculates an evaluation index related to "stability emphasis" and outputs it to the presentation unit 407. The evaluation index related to "stability emphasis" is, for example, at least one of three items: a gain margin and a phase margin calculated based on the open-loop frequency characteristics corresponding to the control parameters related to "stability emphasis", and a control band calculated based on the closed-loop frequency characteristics corresponding to the control parameters related to "stability emphasis".
The control band means the frequency at which the gain crosses 0 dB or −3 dB.

(Presentation unit 407)
The presentation unit 407 acquires the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the initial control parameters before adjustment, the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "standard", and the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "stability emphasis" from the frequency characteristic prediction unit 403. In addition, the presentation unit 407 acquires the evaluation index for "before adjustment", the evaluation index for "standard", and the evaluation index for "stability emphasis" from the evaluation index calculation unit 406.
The presentation unit 407 then creates a Bode diagram from the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the initial control parameters before adjustment, the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "standard", and the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "stability-oriented", and displays them on the display screen together with evaluation indexes for "before adjustment", "standard", and "stability-oriented".
The presentation of the Bode diagram and the evaluation index by the presentation unit 407 is not limited to a display on a display screen, but may be a presentation on paper printed by a printer, etc. In the following description, an example in which the presentation unit 407 displays on a display screen will be described.

7 is a diagram showing a display screen that displays Bode diagrams of closed-loop frequency characteristics for "before adjustment,""standard," and "stability-oriented," and evaluation indexes for "before adjustment,""standard," and "stability-oriented." Although all of the gain margin, phase margin, and control band are displayed in FIG. 7, one or two of the gain margin, phase margin, and control band may be displayed.
In FIG. 7 , a table 501 displayed on a display screen 500 shows Bode diagrams of open-loop frequency characteristics and closed-loop frequency characteristics for “before adjustment,” “standard,” and “stability-oriented” in display fields 502A, 502B, and 502C, respectively, and shows evaluation indexes for “before adjustment,” “standard,” and “stability-oriented” in display fields 503A, 503B, and 503C, respectively.

8, 9, and 10 are Bode diagrams shown in display fields 502A, 502B, and 502C, respectively. In Fig. 8, 9, and 10, the solid lines indicate the closed-loop frequency characteristics, and the dashed lines indicate the open-loop frequency characteristics.
As described above, the gain margin and phase margin are calculated based on the open-loop frequency characteristics, and the control band is calculated based on the closed-loop frequency characteristics. Therefore, when all of the gain margin, phase margin, and control band are displayed in display fields 503A, 503B, and 503C, Bode diagrams showing the open-loop frequency characteristics and the closed-loop frequency characteristics are displayed in display fields 502A, 502B, and 502C, as shown in FIGS.
When only one or both of the gain margin and the phase margin are displayed in the display fields 503A, 503B, and 503C, Bode diagrams showing open-loop frequency characteristics are displayed in the display fields 502A, 502B, and 502C. When only the control band is displayed in the display fields 503A, 503B, and 503C, Bode diagrams showing closed-loop frequency characteristics are displayed in the display fields 502A, 502B, and 502C.

The presentation unit 407 may display either a Bode diagram of the open-loop frequency characteristics and the closed-loop frequency characteristics for “before adjustment,” “standard,” and “stability-oriented,” or an evaluation index for “before adjustment,” “standard,” and “stability-oriented.”
The presentation unit 407 may display the open-loop frequency characteristics and/or the closed-loop frequency characteristics for a plurality of adjustment conditions in a single Bode diagram. Fig. 11 is a Bode diagram that adds the open-loop frequency characteristics and the closed-loop frequency characteristics for "response-oriented" and "custom" to the open-loop frequency characteristics and the closed-loop frequency characteristics for "standard" and "stability-oriented." In Fig. 11, it goes without saying that the open-loop frequency characteristics or the closed-loop frequency characteristics for a plurality of adjustment conditions may be displayed in a single Bode diagram, as necessary.
In this way, by displaying the open-loop frequency characteristics and/or closed-loop frequency characteristics for the four adjustment conditions, "standard,""stability-oriented,""responsiveness-oriented," and "custom," on a single Bode diagram, it becomes easy to compare the open-loop frequency characteristics and/or closed-loop frequency characteristics for the four adjustment conditions.

The presentation unit 407 does not have to display the Bode diagram of the open-loop frequency characteristics and/or closed-loop frequency characteristics for "before adjustment" and the evaluation index for "before adjustment." In this case, the evaluation index calculation unit 406 does not have to calculate the evaluation index for "before adjustment."

(Control parameter setting unit 408)
The user determines the adjustment conditions by looking at the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the initial control parameters "before adjustment", the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "standard" and the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the control parameters for "stability emphasis", and the evaluation indexes for "before adjustment", "standard", and "stability emphasis", which are displayed on the display screen of the presentation unit 407.
The control parameter setting unit 408 displays the setting screen shown in Fig. 12, and the user selects the determined adjustment conditions on the setting screen. The user selects and inputs standard from among the four adjustment conditions shown in Fig. 12, for example, standard, stability-oriented, responsiveness-oriented, and custom. The control parameter setting unit 408 then reads out the control parameters related to "standard" from the control parameter storage unit 405 and sets them as control parameters for the motor control unit 100.

In addition, the user can verify the validity of the control parameters determined by the adjustment unit 400 by operating the motor control unit 100 set to the control parameters determined by the adjustment unit 400 of this embodiment and measuring the frequency characteristics using the frequency characteristic measurement unit 300.

Next, the operation of the adjustment unit 400 in this embodiment will be explained with reference to the flowchart in Figure 13.

In step S11, the frequency characteristic storage unit 401 stores the closed-loop frequency characteristics and open-loop frequency characteristics of input/output gain and phase delay output from the frequency characteristic measurement unit 300.

In step S12, the adjustment condition setting unit 402 displays a setting screen for inputting adjustment conditions, and sets multiple stability margins (gain margin and phase margin) of the open-loop circuit of the motor control unit 100 for each of the multiple adjustment conditions input by the user.

In step S13, the control parameter adjustment unit 404 outputs the control parameter adjustment information to the frequency characteristic prediction unit 403 and proceeds to step S14.

In step S14, the frequency characteristic prediction unit 403 calculates the frequency characteristics C1 and C2 using the control parameters before and after adjustment, as already described, and predicts the open-loop frequency characteristics after adjustment of the control parameters under one of the multiple adjustment conditions based on the frequency characteristics C1 and C2 and the open-loop frequency characteristics Po obtained from the frequency characteristic storage unit 401. After predicting the open-loop frequency characteristics, the closed-loop frequency characteristics are also predicted using the open-loop frequency characteristics.

In step S15, the control parameter adjustment unit 404 determines whether the stability conditions (stability margin, etc.) are satisfied.

In step S<b>16 , the control parameter adjustment unit 404 stores the adjusted and optimized control parameters in the control parameter storage unit 405 .
In step S17, it is determined whether or not there are other adjustment conditions. If there are other adjustment conditions, the process returns to step S13, and if there are no other adjustment conditions, the process proceeds to step S18.
In step S18, the presentation unit 407 displays the frequency characteristics under a plurality of adjustment conditions, and also displays the evaluation index calculated by the evaluation index calculation unit 406.

According to the present embodiment described above, a single measurement of frequency characteristics can determine multiple frequency characteristics when control parameters such as motor gain and filter coefficients are adjusted under multiple adjustment conditions. As a result, by checking multiple frequency characteristics and/or evaluation indices of multiple frequency characteristics, it is possible to easily compare frequency characteristics and/or evaluation indices of frequency characteristics after adjustment under different adjustment conditions, and easily select the control parameters you want to apply.

<Modification in which the control parameter adjustment unit is replaced with a machine learning unit>
FIG. 14 is a block diagram showing a modified example in which the control parameter adjustment unit 404 of the adjustment unit 400 shown in FIG. 3 is replaced with a machine learning unit 600.
The adjustment unit 400A is the same as the adjustment unit 400 shown in FIG. 3 except that a machine learning unit 600 is used in the control parameter adjustment unit 404.
In the following explanation, we will explain the case where the machine learning unit 600 performs reinforcement learning, but the learning performed by the machine learning unit 600 is not limited to reinforcement learning, and the present invention can also be applied to cases where supervised learning is performed, for example.

The machine learning unit 600 performs Q-learning, taking the estimated values of input/output gain and phase delay output from the frequency characteristic prediction unit 403 as state S and adjusting the values of control parameters related to state S as action A. As is well known to those skilled in the art, the purpose of Q-learning is to select, as the optimal action, the action A with the highest value Q(S, A) from among the possible actions A when in a certain state S.

Specifically, the agent (machine learning device) selects various actions A under a certain state S, and learns the correct value Q(S, A) by selecting a better action based on the reward given for that action A.

Furthermore, since we want to maximize the total rewards that will be obtained in the future, we aim to ultimately achieve Q(S, A) = E[Σ(γ ^t )r _t ]. Here, E[ ] represents the expected value, t is time, γ is a parameter called the discount rate (described later), r _t is the reward at time t, and Σ is the total at time t. The expected value in this equation is the expected value when the state changes in accordance with the optimal action. An update equation for such value Q(S, A) can be expressed, for example, by the following equation 6 (hereinafter referred to as equation 6).

In the above equation 6, S _t represents the state of the environment at time t, and A _t represents an action at time t. Action A _t changes the state to S _t+1 . _{r t+1} represents the reward obtained by that change in state. Furthermore, the term with max is the Q value multiplied by γ when action A with the highest Q value known at that time is selected under state S _t+1 . Here, γ is a parameter with a range of 0<γ≦1 and is called the discount rate. Furthermore, α is a learning coefficient with a range of 0<α≦1.

The above-mentioned formula 6 represents a method for updating the value Q(S _t , A _t ) of the action A _t in the state S _t based on the reward r _t+1 returned as a result of the trial A _t .

The machine learning unit 600 observes the state information S including the frequency characteristics of the input/output gain and phase delay for each frequency estimated by the frequency characteristic prediction unit 403, and determines an action A. The machine learning unit 600 receives a reward every time it performs an action A. The reward will be described later.
In Q-learning, the machine learning unit 600 searches, for example, by trial and error, for the optimal action A that maximizes the total reward over the future. In this way, the machine learning unit 600 can select the optimal action A (i.e., the optimal servo parameter value) for the state S.

FIG. 15 is a block diagram showing the configuration of the machine learning unit 600.
15 , in order to perform the above-described reinforcement learning, the machine learning unit 600 includes a state information acquisition unit 601, a learning unit 602, a behavior information output unit 603, a value function storage unit 604, and an optimized behavior information output unit 605. The learning unit 602 includes a reward output unit 6021, a value function update unit 6022, and a behavior information generation unit 6023.

The state information acquisition unit 601 acquires estimated values of the frequency characteristics of the input/output gain and phase delay of the motor control unit 100 calculated using the adjusted control parameters from the frequency characteristic prediction unit 403 and outputs them to the learning unit 602. This state information S corresponds to the environmental state S in Q learning. In addition, the state information acquisition unit 601 acquires image data relating to a complex plane including a circle that forms a closed curve and a unit circle from the adjustment condition setting unit 402 and outputs it to the learning unit 602.

The integral gain K1v and proportional gain K2v of the speed control unit 120 and the coefficients _ωc , τ, and δ of the transfer function of the filter 130 at the time when Q-learning is first started are generated in advance by the user. In this embodiment, the initial set values of the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130 generated by the user are adjusted to optimal values by reinforcement learning.
If the operator has adjusted the machine tool in advance, the integral gain K1v, the proportional gain K2v, and the coefficients ω _c , τ, and δ may be machine-learned using the adjusted values as initial values.

The learning unit 602 is the part that learns the value Q(S, A) when selecting a certain action A under a certain environmental state S.

First, the reward output unit 6021 of the learning unit 602 will be described.
The reward output unit 6021 is a part that calculates a reward when an action A is selected under a certain state S.

The reward output unit 6021 acquires image data relating to a complex plane including a circle forming a closed curve and a unit circle from the state information acquisition unit 601. The reward output unit 6021 creates a Nyquist locus by plotting the open-loop frequency characteristic H(jω) on the acquired complex plane using the input/output gain and phase delay acquired from the state information acquisition unit 601. The method for creating the Nyquist locus has already been explained in the explanation of the operation of the adjustment unit 400, so it will not be repeated here. In this way, a complex plane showing the Nyquist locus, the unit circle, and the circle passing through the gain margin and phase margin, as shown in FIG. 5, is obtained.
The Nyquist locus in the initial state before the control parameters are adjusted can be created by the reward output unit 6021 acquiring the open-loop frequency characteristic H(jω) from the frequency characteristic storage unit 401 via the state information acquisition unit 601 and drawing the open-loop frequency characteristic H(jω) on the complex plane.
The Nyquist locus in the Q-learning process can be created by the reward output unit 6021 drawing the open-loop frequency characteristic H(jω)′ or the closed-loop frequency characteristic G(jω)′ output from the frequency characteristic prediction unit 403 on a complex plane.

In the following explanation, the radius of the circle is referred to as radius r, and the shortest distance between the circle and the Nyquist locus is referred to as shortest distance d. Here, shortest distance d is the shortest distance between the center of the circle and the Nyquist locus, but it is not limited to this and may also be the shortest distance between the circumference of the circle and the Nyquist locus, for example.

The reward output unit 6021 gives a negative reward if the shortest distance d is smaller than the radius r (d<r) and the Nyquist locus passes inside the closed curve. On the other hand, the reward output unit 6021 gives a zero reward if the shortest distance d is equal to or larger than the radius r (d≧r) and the Nyquist locus does not pass inside the circle.

By providing rewards as described above, the reward output unit 6021 searches by trial and error for the integral gain K1v and proportional gain K2v of the speed control unit 120 and the coefficients _ωc , τ, and δ of the transfer function of the filter 130 such that the Nyquist locus does not pass inside the circle and the gain margin and phase margin are equal to or greater than the values set by the user.

In the example described above, whether the Nyquist locus passes inside a circle that is a closed curve is determined based on the shortest distance between the circle and the Nyquist locus, but this method is not limited to this and other methods may be used. For example, it may be determined based on whether the Nyquist locus touches or intersects with the outer periphery of the circle that is a closed curve.

(Example considering response speed)
When the Nyquist locus passes on a circle (d = r) or outside the circle (d > r), the gain margin and phase margin increase as the Nyquist locus moves away from the circle, increasing the stability of the servo system, but decreasing the feedback gain and response speed.
Therefore, it is desirable that the reward output unit 6021 give a reward so that the feedback gain is as large as possible, at or above the gain margin and phase margin determined by the user. Below, three examples of methods for the reward output unit 6021 to determine a reward so that the feedback gain is as large as possible, at or above the gain margin and phase margin determined by the user, will be described.

(1) Method of determining reward based on cutoff frequency The reward output unit 6021 creates a Bode diagram from the input/output gain and phase delay of the motor control unit 100 calculated using the adjusted control parameters output from the frequency characteristic prediction unit 403, and determines the cutoff frequency.
The cutoff frequency is, for example, the frequency at which the gain characteristic of the Bode diagram is −3 dB, or the frequency at which the phase characteristic is −180 degrees.

The reward output unit 6021 determines the reward so that the cutoff frequency becomes larger.
Specifically, the reward output unit 6021 modifies the integral gain K1v and proportional gain K2v, and/or the coefficients _ωc , τ, and δ, and determines the reward based on whether the cutoff frequency fcut increases, remains the same, or decreases when the state changes from the state S before modification to state S'. In the following description, the cutoff frequency fcut in state S will be referred to as fcut(S), and the cutoff frequency fcut in state S' will be referred to as fcut(S').

When the state changes from S to S' and the cutoff frequency fcut increases, the reward output unit 6021 determines that the cutoff frequency fcut(S') is greater than the cutoff frequency fcut(S) and gives a positive reward.
When the state changes from S to S', if the cutoff frequency fcut does not change, the reward output unit 6021 gives a reward of zero, assuming that the cutoff frequency fcut(S')=the cutoff frequency fcut(S).
When the state changes from S to S' and the cutoff frequency fcut becomes smaller, the reward output unit 6021 determines that the cutoff frequency fcut(S') is smaller than the cutoff frequency fcut(S) and gives a negative reward.

By determining the reward as described above, when the Nyquist locus passes on or outside the circle, the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130 are searched for by trial and error so that the cutoff frequency fcut becomes large.
As the cutoff frequency fcut increases, the feedback gain increases and the response speed becomes faster.

(2) Method of determining reward based on closed-loop characteristics The reward output unit 6021 obtains the closed-loop transfer function G(jω) from the input/output gain and phase delay of the motor control unit 100 calculated using the adjusted control parameters output from the frequency characteristic prediction unit 403. The reward output unit 6021 can apply f=Σ|1−G(jω)| ² as the evaluation function f in a preset frequency domain.
The reward output unit 6021 determines the reward so that the value of the evaluation function f becomes small.
Specifically, the reward output unit 6021 modifies the integral gain K1v and proportional gain K2v, and/or the coefficients _ωc , τ, and δ, and determines the reward based on whether the value of the evaluation function f decreases, remains the same, or increases when the state changes from the state S before modification to state S'. In the following explanation, the value of the evaluation function f in state S will be referred to as f(S), and the value of the evaluation function f in state S' will be referred to as f(S').
If the value of the evaluation function f becomes smaller, the cutoff frequency of the closed loop Bode diagram shown in FIG. 11 becomes larger.

When the state changes from S to S' and the value of the evaluation function f becomes smaller, the reward output unit 6021 gives a positive reward, assuming that the value of the evaluation function f(S') is smaller than the value of the evaluation function f(S).
When the state changes from S to S' and the value of the evaluation function f does not change, the reward output unit 6021 gives a reward of zero, assuming that the value of the evaluation function f(S') = the value of the evaluation function f(S).
When the state changes from S to S' and the value of the evaluation function f increases, the reward output unit 6021 gives a negative reward, assuming that the value of the evaluation function f(S') is greater than the value of the evaluation function f(S).

By determining the reward as described above, the integral gain K1v and proportional gain K2v of the speed control unit 120 and the coefficients _ωc , τ, and δ of the transfer function of the filter 130 are searched for by trial and error so that the value of the evaluation function f becomes small when the Nyquist locus passes on or outside the circle.
As the value of the evaluation function f decreases, the feedback gain increases and the response speed increases.

(3) A method of determining the reward so that the shortest distance d approaches the radius r. When the Nyquist locus passes through a circle (d = r) or outside the circle (d > r), the reward is determined so that the Nyquist locus approaches a closed curve.
Specifically, the reward output unit 6021 modifies the integral gain K1v and proportional gain K2v, and/or the coefficients _ωc , τ, and δ, and determines the reward based on whether the shortest distance d between the center of the circle and the Nyquist locus becomes smaller, the same, or larger when the state changes from the state S before the modification to state S'. In the following description, the shortest distance d in state S will be referred to as d(s), and the shortest distance d in state S' will be referred to as d(s').

When the state changes from S to S' and the shortest distance d becomes small, the reward output unit 6021 determines that shortest distance d(S')<shortest distance d(S) and gives a positive reward.
When the state changes from S to S', if the shortest distance d remains unchanged, the reward output unit 6021 gives a reward of zero, assuming that the shortest distance d(S')=the shortest distance d(S).
When the state changes from S to S', if the shortest distance d becomes large, the reward output unit 6021 determines that shortest distance d(S')>shortest distance d(S) and gives a negative reward.

By determining the reward as described above, the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients ω _c , τ, and δ of the transfer function of the filter 130 are searched for by trial and error so that the Nyquist locus passes through a circle or approaches the outer periphery of the circle.
As the Nyquist locus passes through a circle or approaches the outer periphery of the circle, the feedback gain increases and the response speed becomes faster.
The method of determining the reward based on the information of the shortest distance d is not limited to the above method, and other methods can be applied.

(Example taking resonance into consideration)
Even when the Nyquist locus passes on the circle (d=r) or outside the circle (d>r), the input/output gain may increase due to resonance at the machine end of the machine to be controlled.
Therefore, it is desirable that the reward output unit 6021 determines the reward so as to suppress resonance at or above the gain margin and phase margin determined by the user. Below, a method for determining the reward by comparing the open loop characteristics with a reference model will be described.

Below, using Figures 16 and 17, we will explain the operation of the reward output unit 6021 to give a negative reward when the input/output gain for each frequency in the created frequency characteristic is greater than the input/output gain of the reference model.

The reward output unit 6021 stores a reference model of input/output gain. The reference model is a model of a motor control unit that has ideal characteristics without resonance. The reference model can be calculated from the inertia Ja, torque constant _Kt , proportional gain _Kp , integral gain _KI , and derivative gain _KD of the model shown in FIG. 16, for example. The inertia Ja is the sum of the motor inertia and the machine inertia.

FIG. 17 is a characteristic diagram showing the frequency characteristics of the input/output gain between the motor control unit of the reference model and the motor control unit 100 before and after learning. As shown in the characteristic diagram of FIG. 17, the reference model has a frequency region FA where the ideal input/output gain is equal to or greater than a certain input/output gain, for example, -20 dB, and a frequency region FB where the ideal input/output gain is less than the certain input/output gain. In region FA of FIG. 17, the ideal input/output gain of the reference model is shown by curve _MC1 (thick line). In region FB of FIG. 17, the ideal virtual input/output gain of the reference model is shown by curve _MC11 (thick dashed line), and the input/output gain of the reference model is a constant value shown by line _MC12 (thick line). In regions FA and FB of FIG. 13, the curves of the input/output gain with the motor control unit before and after learning are shown by curves _RC1 and _RC2 , respectively.

In the area FA, the reward output unit 6021 gives a negative reward when the curve _RC1 of the input/output gain for each frequency in the created frequency characteristic before learning exceeds the curve _MC1 of the ideal input/output gain of the reference model.
In the region FB exceeding the frequency where the input/output gain becomes sufficiently small, the impact on stability is small even if the pre-learning input/output gain curve _RC1 exceeds the ideal virtual input/output gain curve _MC11 of the reference model. Therefore, in the region FB, as described above, the input/output gain of the reference model uses the straight line _MC12 of a constant input/output gain (for example, -20 dB) rather than the ideal gain characteristic curve _MC11 . However, if the measured input/output gain curve _RC1 before learning exceeds the straight line _MC12 of a constant input/output gain, there is a possibility of instability, so a negative value is given as a reward.

When adjusting the input/output gain, the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130 are adjusted. The characteristics of the filter 130 vary in gain and phase depending on the bandwidth fw of the filter 130, and vary in gain and phase depending on the attenuation coefficient k of the filter 130. Therefore, the input/output gain can be adjusted by adjusting the coefficients of the filter 130.

If the reward output unit 6021 assigns a negative reward when the shortest distance d is smaller than the radius r (d<r) and the Nyquist locus passes inside the closed curve, it outputs this negative reward to the value function update unit 6022. If the reward output unit 6021 assigns a positive reward when the shortest distance d is equal to or larger than the radius r (d≧r) and the Nyquist locus does not pass inside the circle, it outputs this positive reward to the value function update unit 6022.
When reward output unit 6021 has given a reward in the three examples taking response speed into consideration or the example taking resonance into consideration, it outputs the total reward obtained by adding this reward to the positive value reward that is given when the Nyquist locus does not pass through the inside of the circle to value function update unit 6022.

When adding rewards, weights may be applied to the rewards. For example, when emphasis is placed on the stability of the servo system, a positive reward given when the Nyquist locus does not pass through the inside of the circle can be weighted to be more important than the rewards given in the three examples taking response speed into consideration or the example taking resonance into consideration.
The reward output unit 6021 has been described above.

The value function update unit 6022 updates the value function Q stored in the value function memory unit 604 by performing Q-learning based on the state S, the action A, the state S′ when the action A is applied to the state S, and the reward calculated as described above.
The value function Q may be updated by online learning, batch learning, or mini-batch learning.
Online learning is a learning method in which a certain action A is applied to the current state S, and the value function Q is updated immediately each time the state S transitions to a new state S'. Batch learning is a learning method in which a certain action A is applied to the current state S, and the state S transitions to a new state S', thereby collecting learning data and updating the value function Q using all of the collected learning data. Mini-batch learning is a learning method that is intermediate between online learning and batch learning, and updates the value function Q each time a certain amount of learning data is accumulated.

The behavior information generation unit 6023 selects behavior A in the Q-learning process for the current state S. The behavior information generation unit 6023 generates behavior information A and outputs the generated behavior information A to the behavior information output unit 603 in order to perform an operation (corresponding to behavior A in Q-learning) to modify the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130 in the Q-learning process.
More specifically, the behavior information generation unit 6023 may, for example, incrementally add or subtract the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the respective coefficients _ωc , τ, and δ of the transfer function of the filter 130 included in behavior A to the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the respective coefficients _ωc , τ, and δ of the transfer function of the filter 130 included in state S.

Note that the integral gain K1v and proportional gain K2v of the speed control unit 120 and the coefficients _ωc , _{τ, and δ of the filter 130 may all be modified, or some of the coefficients may be modified. When the coefficients ωc} , τ, and δ of the filter 130 are modified, for example, the center frequency fc at which resonance occurs is easy to find and identify. Therefore, the behavior information generation unit 6023 may temporarily fix the center frequency fc and modify the bandwidth fw and the attenuation coefficient δ, that is, may fix the coefficient _ωc (= 2πfc) and modify the coefficient τ (= fw/fc) and the attenuation coefficient δ, and may generate behavior information A and output the generated behavior information A to the behavior information output unit 603.

In addition, the behavioral information generation unit 6023 may take a strategy of selecting behavior A' using a known method, such as a greedy method that selects behavior A' with the highest value Q(S, A) among the currently estimated values of behavior A, or an ε-greedy method that randomly selects behavior A' with a certain small probability ε and otherwise selects behavior A' with the highest value Q(S, A).

The behavioral information output unit 603 is a part that transmits the behavioral information A output from the learning unit 602 to the frequency characteristic prediction unit 403. As described above, the filter 130 makes a transition to the next state S' (i.e., the corrected integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the respective coefficients _ωc , τ, and δ) based on this behavioral information by finely correcting the current state S, i.e., the currently set integral gain K1v and proportional gain K2v of the speed control unit 120, and/or the respective coefficients ωc, τ, and δ.

The value function memory unit 604 is a storage device that stores the value function Q. The value function Q may be stored as a table (hereinafter referred to as an action value table) for each state S and action A, for example. The value function Q stored in the value function memory unit 604 is updated by the value function update unit 6022. The value function Q stored in the value function memory unit 604 may also be shared with other machine learning units 600. If the value function Q is shared by multiple machine learning units 600, reinforcement learning can be performed in a distributed manner among the machine learning units 600, thereby improving the efficiency of reinforcement learning.

The optimization behavior information output unit 605 generates behavior information A (hereinafter referred to as "optimization behavior information") for causing the speed control unit 120 and the filter 130 to perform an operation that maximizes the value Q(S, A) based on the value function Q updated by the value function update unit 6022 through Q learning.
More specifically, the optimization behavior information output unit 605 acquires the value function Q stored in the value function storage unit 604. This value function Q has been updated by the value function update unit 6022 performing Q-learning, as described above. The optimization behavior information output unit 605 then generates behavior information based on the value function Q and outputs the generated behavior information to the control parameter storage unit 405. Similar to the behavior information output by the behavior information output unit 603 in the Q-learning process, this optimization behavior information includes information for modifying the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130.
Through the above operations, the machine learning unit 600 can optimize the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130, and operate the motor control unit 100 so that the stability margin is equal to or greater than a predetermined value.
Furthermore, the above operation optimizes the integral gain K1v and proportional gain K2v of the speed control unit 120 and/or the coefficients _ωc , τ, and δ of the transfer function of the filter 130, thereby making the stability margin of the motor control unit 100 equal to or greater than a predetermined value, and increasing the feedback gain to increase the response speed and/or suppress resonance.
As described above, by using the machine learning unit 600 of the present disclosure, it is possible to simplify the adjustment of the gain of the speed control unit 120 and the parameters of the filter 130.

Second Embodiment
FIG. 18 is a block diagram showing a configuration example of an adjustment unit included in the control system according to the second embodiment of the present disclosure.
The adjustment unit 400B shown in FIG. 18 differs from the adjustment unit 400 shown in FIG. 3 in that a time response prediction unit 409 and an evaluation index calculation unit 410 are provided.
The time response prediction unit 409, the evaluation index calculation unit 410, the presentation unit 407, and the control parameter setting unit 408 will be described below.

(Time response prediction unit 409)
The time response prediction unit 409 acquires the open-loop frequency characteristics/closed-loop frequency characteristics of the initial state before adjustment from the frequency characteristic prediction unit 403, and predicts the time response before adjustment (which becomes the second time response). The time response prediction unit 409 also acquires the open-loop frequency characteristics/closed-loop frequency characteristics corresponding to the control parameters optimized for "standard," and predicts the time response corresponding to the "standard" adjustment condition (which becomes the first time response). The time response prediction unit 409 also acquires the open-loop frequency characteristics/closed-loop frequency characteristics corresponding to the control parameters optimized for "stability-oriented," and predicts the time response corresponding to the "stability-oriented" adjustment condition (which becomes the first time response).
The method of predicting time response using frequency characteristics involves performing modal analysis using information on frequency characteristics to create a transfer function model P(s). By performing an inverse Laplace transform on this transfer function model P(s), a time domain model y(t) is obtained. The method of predicting time response using frequency characteristics is described in, for example, Japanese Patent No. 6515844.

An example of a method for predicting a time response using frequency characteristics will be described below.
The time response prediction unit 409 acquires the open-loop frequency characteristics and/or closed-loop frequency characteristics corresponding to the optimized control parameters for the "standard" and performs modal analysis. Modal analysis involves estimating the modal frequency ω and modal damping ratio ζ of the mechanical vibration from the frequency characteristics.
For example, a model P(s) of the transfer function of Equation 7 (hereinafter Equation 7) is created by modal analysis. The first term on the right side of Equation 7 is the rigid body mode, and the second term is the resonance mode. _ωn and _ζn indicate the frequency and damping ratio of the nth mode. _K0 and _Kn are coefficients.

Next, principal component analysis is performed to obtain a model P(s) of the transfer function of Equation 8 (Equation 8 below). Principal component analysis extracts only the main (dominant) mode from among the multiple modes obtained from the modal analysis.



The above formulas 7 and 8 are machine models when only the stiffness mode and the first resonance mode are considered, and therefore the characteristics of the machine can be expressed by a model with the minimum necessary degrees of freedom (modes).
FIG. 19 is a Bode diagram showing the first and second resonance modes, and Equations 7 and 8 are examples taking the first resonance mode into consideration.
Furthermore, by performing an inverse Laplace transform on the above equation 8, a time domain model y(t) is obtained.

(Evaluation index calculation unit 410)
The evaluation index calculation unit 410 calculates at least one evaluation index (evaluation index related to "before adjustment") from among the rise time, the amount of overshoot, the settling time, etc., based on the time response corresponding to "before adjustment", and outputs the calculated result to the presentation unit 407. The evaluation index calculation unit 410 also calculates at least one evaluation index (evaluation index related to "standard") from among the rise time, the amount of overshoot, the settling time, etc., based on the time response corresponding to the "standard" adjustment condition, and outputs the calculated result to the presentation unit 407. The evaluation index calculation unit 410 also calculates at least one evaluation index (evaluation index related to "stability emphasis") from among the rise time, the amount of overshoot, the settling time, etc., based on the time response corresponding to the "stability emphasis" adjustment condition, and outputs the calculated result to the presentation unit 407.

(Presentation unit 407)
The presentation unit 407 displays a display screen shown in FIG. 20, which will be described below, in addition to the display screen shown in FIG.
The presentation unit 407 acquires the "before adjustment" time response, the time response corresponding to the "standard" adjustment condition, and the time response corresponding to the "stability-oriented" adjustment condition from the time response prediction unit 409. The presentation unit 407 also acquires the evaluation index related to the "before adjustment" time response, the evaluation index related to the "standard" adjustment condition, and the evaluation index related to the time response corresponding to the "stability-oriented" adjustment condition from the evaluation index calculation unit 406. The time response is, for example, a step response or an impulse response.
The presentation unit 407 then creates characteristic diagrams of the time response from the time response before adjustment, the time response corresponding to the "standard" adjustment condition, and the time response corresponding to the "stability-emphasis" adjustment condition, and displays these on the display screen 700 together with an evaluation index for the time response "before adjustment," an evaluation index for the time response corresponding to the "standard" adjustment condition, and an evaluation index for the time response corresponding to the "stability-emphasis" adjustment condition.
The presentation unit 407 may display either a characteristic diagram of the time response corresponding to "before adjustment,""standard," and "stability-oriented," or an evaluation index for "before adjustment,""standard," and "stability-oriented."

20 is a diagram showing a display screen displaying characteristic diagrams of time responses corresponding to "before adjustment,""standard," and "stability-oriented," as well as evaluation indices related to time responses corresponding to "before adjustment,""standard," and "stability-oriented." While all of the rise time, overshoot amount, and settling time are displayed in FIG. 20, one or two of the rise time, overshoot amount, and settling time may be displayed.
In FIG. 20 , in a table 701 displayed on a display screen 700, display fields 702A, 702B, and 702C show characteristic diagrams of time responses corresponding to “before adjustment,” “standard,” and “stability-oriented,” respectively, and display fields 703A, 703B, and 703C show evaluation indices related to time responses corresponding to “before adjustment,” “standard,” and “stability-oriented,” respectively.
21, 22, and 23 are diagrams showing the time response characteristics shown in display fields 702A, 702B, and 702C, respectively.
The presentation unit 407 may display the time responses for a plurality of adjustment conditions in one characteristic diagram.
The presentation unit 407 may display the display screen shown in Fig. 20 alone or together with the display screen shown in Fig. 7. Furthermore, the presentation unit 407 may be configured with a presentation unit that displays the display screen shown in Fig. 20 and a presentation unit that displays the display screen shown in Fig. 7.
In this embodiment, if the display screen shown in FIG. 7 is not displayed and only the display screen shown in FIG. 20 is displayed, the evaluation index calculation unit 406 does not need to be provided, and the closed-loop frequency characteristics and/or the open-loop frequency characteristics do not need to be input from the frequency characteristic prediction unit 403 to the presentation unit 407.

The presentation unit 407 does not need to display the time response "before adjustment" or the evaluation index for the time response "before adjustment." In this case, the evaluation index calculation unit 410 does not need to calculate the evaluation index for the time response "before adjustment."

(Control parameter setting unit 408)
The user determines the adjustment conditions by looking at the time response characteristic diagram before adjustment, the time response characteristic diagram corresponding to the "standard" adjustment condition, and the time response characteristic diagram corresponding to the "stability-emphasis" adjustment condition, all displayed on the display screen of the presentation unit 407, as well as the evaluation indexes related to the time response corresponding to "before adjustment,""standard," and "stability-emphasis."
The control parameter setting unit 408 displays the setting screen shown in Fig. 12, and the user selects adjustment conditions by viewing this setting screen. For example, the user selects "standard" from among four adjustment conditions: standard, emphasis on stability, emphasis on responsiveness, and custom, and inputs the selected condition. The control parameter setting unit 408 then reads out the control parameters for "standard" from the control parameter storage unit 405 and sets them as control parameters for the motor control unit 100.
When the presentation unit 407 displays the display screen shown in FIG. 7 together with the display screen shown in FIG. 20, the user can view both display screens and select an adjustment condition.

The control system 10 and the functional blocks included in the adjustment units 400, 400A, and 400B have been described above.
To realize these functional blocks, the control system 10 or any of the adjustment units 400, 400A, and 400B includes a processing unit such as a CPU (Central Processing Unit). The control system 10 or any of the adjustment units 400, 400A, and 400B also includes a main storage device such as an auxiliary storage device such as an HDD (Hard Disk Drive) that stores various control programs such as application software or an OS (Operating System), and a RAM (Random Access Memory) that stores data temporarily required for the processing unit to execute the programs.

Then, in the control system 10 or in any of the adjustment units 400, 400A, and 400B, the arithmetic processing unit loads the application software or OS from the auxiliary storage device, and while expanding the loaded application software or OS into the main storage device, performs arithmetic processing based on the application software or OS. Furthermore, based on the results of this calculation, it controls the various hardware components of each device. This is how the functional blocks of this embodiment are realized. In other words, this embodiment can be realized through the cooperation of hardware and software.

As the machine learning unit 600 requires a large amount of computation for machine learning, it is advantageous to install a GPU (Graphics Processing Unit) in a personal computer and use the GPU for the computational processing required for machine learning using a technology known as GPGPU (General-Purpose Computing on Graphics Processing Units), thereby enabling high-speed processing. Furthermore, to achieve even faster processing, a computer cluster may be constructed using multiple computers equipped with such GPUs, and parallel processing may be performed on the multiple computers included in this computer cluster.

The above-mentioned control system 10 and each component included in adjustment units 400, 400A, and 400B can be realized by hardware, software, or a combination of these. Furthermore, the control parameter adjustment method performed by the cooperation of each component included in control system 10 and adjustment units 400, 400A, and 400B can also be realized by hardware, software, or a combination of these. Here, "realized by software" means realized by a computer reading and executing a program.

The program can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R/Ws, and semiconductor memories (e.g., mask ROMs, programmable ROMs (PROMs), erasable PROMs (EPROMs), flash ROMs, and random access memory (RAMs)). The program may also be supplied to a computer by various types of transitory computer-readable media.

The above-described embodiment is a preferred embodiment of the present invention, but the scope of the present invention is not limited to the above-described embodiment alone, and the present invention can be implemented in various modified forms within the scope that does not deviate from the gist of the present invention.

In the above-described embodiment, a case where one filter is provided has been described, but the filter 130 may also be configured by connecting multiple filters in series, each corresponding to a different frequency band. Fig. 24 is a block diagram showing an example of a filter configured by directly connecting multiple filters. In Fig. 24, when there are m resonance points (m is a natural number greater than or equal to 2), the filter 130 is configured by connecting m filters 130-1 to 130-m in series. Optimal values for the coefficients _ωc , τ, and δ of each of the m filters 130-1 to 130-m are found by machine learning.

In addition to the configuration shown in FIG. 1, the control system may have the following configurations.
<Modification in which the adjustment unit is provided outside the motor control unit via a network>
Fig. 25 is a block diagram showing another example of the configuration of a control system. Control system 10A shown in Fig. 25 differs from control system 10 shown in Fig. 1 in that n (n is a natural number of 2 or more) motor control units 100-1 to 100-n are connected to n adjustment units 400-1 to 400-n via network 800, and each includes a frequency generation unit 200 and a frequency characteristic measurement unit 300.
Adjustment units 400-1 to 400-n have the same configuration as adjustment units 400, 400A, or 400B. Motor control units 100-1 to 100-n correspond to motor control devices, and adjustment units 400-1 to 400-n correspond to adjustment devices. Of course, one or both of frequency generation unit 200 and frequency characteristic measurement unit 300 may be provided outside motor control units 100-1 to 100-n.

Here, motor control unit 100-1 and adjustment unit 400-1 are paired one-to-one and connected to enable communication. Motor control units 100-2 to 100-n and adjustment units 400-2 to 400-n are connected in the same way as motor control unit 100-1 and adjustment unit 400-1. In FIG. 25, n pairs of motor control units 100-1 to 100-n and adjustment units 400-1 to 400-n are connected via network 800, but the motor control unit and adjustment unit of each of the n pairs of motor control units 100-1 to 100-n and adjustment units 400-1 to 400-n may be directly connected via a connection interface. These n pairs of motor control units 100-1 to 100-n and adjustment units 400-1 to 400-n may be installed in the same factory, for example, or in different factories.

The network 800 may be, for example, a LAN (Local Area Network) established within a factory, the Internet, a public telephone network, or a combination of these. There are no particular limitations on the specific communication method used in the network 800, or whether it is a wired or wireless connection.

<Flexibility of system configuration>
In the above-described embodiment, the motor control units 100-1 to 100-n and the adjustment units 400-1 to 400-n are connected in one-to-one pairs so that they can communicate with each other, but for example, one adjustment unit may be connected to multiple motor control units so that they can communicate with each other via the network 800.
In this case, the functions of one adjustment unit may be distributed to multiple servers as needed to form a distributed processing system.Furthermore, the functions of one adjustment unit may be realized using a virtual server function on the cloud.

Furthermore, if there are n motor control units 100-1 to 100-n of the same model name, same specifications, or same series, and n corresponding adjustment units 400-1 to 400-n, the adjustment results of each adjustment unit 400-1 to 400-n can be configured to be shared. This makes it possible to build a more optimal model.

The adjustment device, control system, and control parameter adjustment method for adjusting a control parameter according to the present disclosure can take various forms including the above-described embodiments, having the following configurations.
(1) An adjustment device (e.g., adjustment unit 400, 400A) that adjusts control parameters of a motor control unit that controls a motor,
a frequency characteristic storage unit (e.g., frequency characteristic storage unit 401) that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit (e.g., an adjustment condition setting unit 402) that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit (for example, a frequency characteristic prediction unit 403) that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit (e.g., a control parameter adjustment unit 404) that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit (e.g., a control parameter storage unit 405) for storing the plurality of control parameters optimized for the plurality of adjustment conditions;
an evaluation index calculation unit (e.g., evaluation index calculation unit 406) that calculates an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameter;
a presentation unit (e.g., presentation unit 407) that presents at least one of the predicted frequency characteristics and the evaluation index corresponding to the optimized control parameters for each of the plurality of adjustment conditions;
a control parameter setting unit (e.g., a control parameter setting unit 408) that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
An adjustment device comprising:
With this adjustment device, a single measurement of frequency characteristics can determine multiple frequency characteristics when control parameters such as the gain of the motor control unit and the coefficient of the filter are adjusted under multiple adjustment conditions. As a result, by checking the multiple frequency characteristics and/or the evaluation indexes of the multiple frequency characteristics, it is possible to easily compare the frequency characteristics and/or the evaluation indexes of the frequency characteristics after adjustment under different adjustment conditions, and easily select the control parameters you want to apply.

(2) An adjustment device (e.g., adjustment unit 400B) that adjusts control parameters of a motor control unit that controls a motor,
a frequency characteristic storage unit (e.g., frequency characteristic storage unit 401) that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit (e.g., an adjustment condition setting unit 402) that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit (for example, a frequency characteristic prediction unit 403) that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit (e.g., control parameter adjustment unit 404) that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit (e.g., a control parameter storage unit 405) for storing the plurality of control parameters optimized for the plurality of adjustment conditions;
a time response prediction unit (e.g., the time response prediction unit 409) that predicts a first time response using predicted frequency characteristics corresponding to the optimized control parameters;
an evaluation index calculation unit (e.g., evaluation index calculation unit 410) that calculates an evaluation index of the first time response from the predicted first time response;
a presentation unit (e.g., presentation unit 407) that presents at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
a control parameter setting unit (e.g., a control parameter setting unit 408) that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
An adjusting device comprising:
This adjustment device can obtain multiple frequency characteristics when control parameters such as the gain of the motor control unit and the coefficient of the filter are adjusted under multiple adjustment conditions by measuring the frequency characteristics once. As a result, by checking multiple time responses and/or multiple evaluation indices of the time responses predicted from the multiple frequency characteristics, it is possible to easily compare the time responses and/or evaluation indices of the time responses after adjustment under different adjustment conditions and easily select the control parameters to be applied.

(3) The evaluation index calculation unit calculates an evaluation index of the frequency characteristics of the measured machine from the frequency characteristics of the measured machine,
The adjustment device according to (1) above, wherein the presentation unit presents at least one of the measured frequency characteristics of the machine and an evaluation index of the measured frequency characteristics of the machine.

(4) The time response prediction unit predicts a second time response using the measured frequency characteristics of the machine,
the evaluation index calculation unit calculates an evaluation index of the second time response from the second time response;
The adjustment device according to (2), wherein the presentation unit presents at least one of the second time response and an evaluation index of the second time response.

(5) An adjustment device described in any of (1) to (4) above, in which the control parameter adjustment unit optimizes the control parameters using machine learning.

(6) An adjustment device described in any of (1) to (5) above, wherein the control parameter is at least one of the gain and filter coefficient of the motor control unit.

(7) An adjustment device described in (1) or (3) above, wherein the evaluation index is at least one of a gain margin, a phase margin, and a control band.

(8) An adjustment device described in (2) or (4) above, wherein the time response is a step response or an impulse response.

(9) An adjustment device described in any one of (2), (4) or (8) above, wherein the evaluation index of the time response is at least one of rise time, overshoot amount, and settling time.

(10) a motor control unit (e.g., motor control unit 100) that controls a motor;
An adjustment device according to any one of (1) to (9) above, which adjusts control parameters of the motor control unit;
A control system with
This control system can obtain multiple frequency characteristics when adjusting control parameters such as the gain of the motor control unit and the coefficients of the filter under multiple adjustment conditions by measuring the frequency characteristics once. As a result, by checking the multiple frequency characteristics and/or the evaluation indexes of the multiple frequency characteristics, it is possible to easily compare the frequency characteristics and/or the evaluation indexes of the frequency characteristics after adjustment under different adjustment conditions and easily select the control parameters to be applied. Alternatively, by checking the multiple time responses and/or the evaluation indexes of the multiple time responses predicted from the multiple frequency characteristics, it is possible to easily compare the time responses and/or the evaluation indexes of the time responses after adjustment under different adjustment conditions and easily select the control parameters to be applied.

(11) A frequency generating unit that generates a signal whose frequency changes and inputs the signal to the motor control unit;
a frequency characteristic measurement unit that measures the frequency characteristics of the machine by measuring the frequency characteristics of the input/output gain and phase delay of the motor control unit based on the signal and an output signal of the motor control unit;
The control system according to (10) above, comprising:

(12) A control parameter adjustment method for adjusting control parameters of a motor control unit (e.g., motor control unit 100) that controls a motor, comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
A process of calculating an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameters;
a process of presenting at least one of predicted frequency characteristics corresponding to the optimized control parameters and an evaluation index of the frequency characteristics for each of a plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
A control parameter tuning method is performed.
According to this control parameter adjustment method, a single measurement of frequency characteristics can determine multiple frequency characteristics when control parameters such as the gain of the motor control unit and the coefficient of the filter are adjusted under multiple adjustment conditions. As a result, by checking the multiple frequency characteristics and/or the evaluation indexes of the multiple frequency characteristics, it is possible to easily compare the frequency characteristics and/or the evaluation indexes of the frequency characteristics after adjustment under different adjustment conditions, and easily select the control parameters to be applied.

(13) A control parameter adjustment method for adjusting control parameters of a motor control unit (e.g., motor control unit 100) that controls a motor, comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
predicting a first time response using the predicted frequency characteristics corresponding to the optimized control parameters;
calculating an evaluation index of the first time response from the predicted first time response;
a process of presenting at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
A control parameter tuning method is performed.
According to this control parameter adjustment method, a single measurement of frequency characteristics can determine multiple frequency characteristics when control parameters such as the gain of the motor control unit and the coefficient of the filter are adjusted under multiple adjustment conditions. As a result, by checking multiple time responses and/or multiple evaluation indices of the time responses predicted from the multiple frequency characteristics, it is possible to easily compare the time responses and/or evaluation indices of the time responses after adjustment under different adjustment conditions, and easily select the control parameters to be applied.

(14) The computer
A process of calculating an evaluation index of the frequency characteristics of the machine measured from the frequency characteristics of the machine measured;
a process of presenting at least one of the measured machine frequency characteristics and an evaluation index of the measured machine frequency characteristics;
The control parameter adjustment method according to (12) above,

(15) The computer
predicting a second time response using the measured frequency characteristics of the machine;
A process of calculating an evaluation index of the second time response from the second time response;
presenting at least one of the second time response and a metric of the second time response;
The control parameter adjustment method according to (13) above,

10, 10A Control system 100 Motor control unit 110 Subtractor 120 Speed control unit 130 Filter 140 Current control unit 150 Motor 200 Frequency generation unit 300 Frequency characteristic measurement unit 400, 400A, 400B Adjustment unit 401 Frequency characteristic storage unit 402 Adjustment condition setting unit 403 Frequency characteristic prediction unit 404 Control parameter adjustment unit 405 Control parameter storage unit 406 Evaluation index calculation unit 407 Presentation unit 408 Control parameter setting unit 409 Time response prediction unit 410 Evaluation index calculation unit 500 Display screen 600 Machine learning unit 601 State information acquisition unit 602 Learning unit 603 Action information output unit 604 Value function storage unit 605 Optimization action information output unit 700 Display screen 800 Network

Claims (15)

An adjustment device that adjusts control parameters of a motor control unit that controls a motor,
a frequency characteristic storage unit that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit that stores the plurality of control parameters optimized for the plurality of adjustment conditions;
an evaluation index calculation unit that calculates an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameter;
a presentation unit that presents at least one of the predicted frequency characteristics and the evaluation index corresponding to the optimized control parameters for each of a plurality of adjustment conditions;
a control parameter setting unit that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
An adjusting device comprising: An adjustment device that adjusts control parameters of a motor control unit that controls a motor,
a frequency characteristic storage unit that stores the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
an adjustment condition setting unit that sets a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a frequency characteristic prediction unit that predicts the frequency characteristics of the machine after the control parameters have been adjusted, using the control parameters before and after the adjustment and the frequency characteristics stored in the frequency characteristic storage unit;
a control parameter adjustment unit that adjusts the control parameter to be input to the frequency characteristic prediction unit in order to optimize the control parameter, using the predicted frequency characteristic and one of a plurality of adjustment conditions set by the adjustment condition setting unit;
a control parameter storage unit that stores the plurality of control parameters optimized for the plurality of adjustment conditions;
a time response prediction unit that predicts a first time response using a predicted frequency characteristic corresponding to the optimized control parameter;
an evaluation index calculation unit that calculates an evaluation index of the first time response from the predicted first time response;
a presentation unit that presents at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
a control parameter setting unit that sets a control parameter selected from the plurality of control parameters stored in the control parameter storage unit to the motor control unit;
An adjustment device comprising: the evaluation index calculation unit calculates an evaluation index of the measured frequency characteristics of the machine from the measured frequency characteristics of the machine,
The adjustment device according to claim 1 , wherein the presentation unit presents at least one of the measured frequency characteristics of the machine and an evaluation index of the measured frequency characteristics of the machine. the time response prediction unit predicts a second time response using the measured frequency characteristics of the machine;
the evaluation index calculation unit calculates an evaluation index of the second time response from the second time response;
The adjustment device according to claim 2 , wherein the presentation unit presents at least one of the second time response and an evaluation index of the second time response. An adjustment device described in any one of claims 1 to 4, wherein the control parameter adjustment unit optimizes the control parameters using machine learning. An adjustment device according to any one of claims 1 to 5, wherein the control parameter is at least one of the gain and filter coefficient of the motor control unit. The adjustment device described in claim 1 or 3, wherein the evaluation index is at least one of a gain margin, a phase margin, and a control band. The adjustment device described in claim 2 or 4, wherein the time response is a step response or an impulse response. The adjustment device described in claim 2, 4, or 8, wherein the evaluation index of the time response is at least one of rise time, overshoot amount, and settling time. a motor control unit that controls the motor;
an adjustment device according to claim 1 that adjusts control parameters of the motor control unit;
A control system with a frequency generating unit that generates a signal whose frequency changes and inputs the signal to the motor control unit;
a frequency characteristic measurement unit that measures the frequency characteristics of the machine by measuring the frequency characteristics of the input/output gain and phase delay of the motor control unit based on the signal and an output signal of the motor control unit;
The control system of claim 10, comprising: A control parameter adjustment method for adjusting control parameters of a motor control unit that controls a motor, comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
A process of calculating an evaluation index of the frequency characteristic from the predicted frequency characteristic corresponding to the optimized control parameters;
a process of presenting at least one of the predicted frequency characteristics and the evaluation index corresponding to the optimized control parameters for each of the plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
A control parameter tuning method is performed. A control parameter adjustment method for adjusting control parameters of a motor control unit that controls a motor, comprising:
The computer
A process of storing the frequency characteristics of the machine measured by operating the motor control unit having the control parameters before adjustment;
a process of setting a plurality of adjustment conditions for adjusting the control parameters of the motor control unit;
a process of predicting the frequency characteristics of the machine after adjusting the control parameters, using the control parameters before and after adjustment and the stored frequency characteristics;
a process of adjusting the control parameters to optimize them using the predicted frequency characteristics and one of a plurality of set adjustment conditions;
a process of storing the plurality of control parameters optimized for the plurality of adjustment conditions;
predicting a first time response using the predicted frequency characteristics corresponding to the optimized control parameters;
calculating an evaluation index of the first time response from the predicted first time response;
a process of presenting at least one of the first time response and the evaluation index for each of a plurality of adjustment conditions;
A process of setting a control parameter selected from the plurality of stored control parameters to the motor control unit;
A control parameter tuning method is performed. The computer
A process of calculating an evaluation index of the frequency characteristics of the machine measured from the frequency characteristics of the machine measured;
a process of presenting at least one of the measured machine frequency characteristics and an evaluation index of the measured machine frequency characteristics;
The control parameter adjusting method according to claim 12, further comprising the steps of: The computer
predicting a second time response using the measured frequency characteristics of the machine;
A process of calculating an evaluation index of the second time response from the second time response;
presenting at least one of the second time response and a metric of the second time response;
The control parameter adjusting method according to claim 13, further comprising the steps of: