JP4396541B2 - Control parameter sensitivity analysis device for motor control device and control parameter setting method for motor control device - Google Patents

Description

  The present invention relates to a motor control device used in a positioning device such as a semiconductor manufacturing device or a machine tool or an industrial robot, and is intended to optimally adjust a control parameter of the motor control device. The present invention relates to a control parameter sensitivity analysis device for a motor control device and a control parameter setting method for the motor control device for confirming the result.

  Conventionally, a control parameter sensitivity analysis device for a motor control device for confirming a control parameter adjustment result for the purpose of optimally adjusting a control parameter of the motor control device is disclosed in, for example, Patent Literature 1 and Patent Literature 2. There is something that was done.

  First, Patent Document 1 as a first prior art discloses an exposure apparatus, a vibration isolation apparatus, and a system proposed for the purpose of capturing and identifying changes over time in dynamic characteristics of a motion mechanism with high accuracy in a short time. An identification apparatus and method for a sensor that applies a pseudo-random signal to an actuator for controlling the motion in a motion mechanism and measures the motion state of the motion mechanism together with time-series data of the pseudo-random signal. The time series data is collected and stored in the data storage means, the time series data stored in the data storage means is filtered, the system identification is performed using the filtered data, and the mathematics calculated by the system identification is calculated. Deriving characteristic values of the motion mechanism using a model and capturing changes in the characteristic values over time to self-diagnose the suitability of maintaining the performance of the motion mechanism Going on. Specific examples are shown below.

FIG. 18 is a block diagram showing a system identification apparatus and method applied to an active vibration isolator in a semiconductor manufacturing apparatus according to the first prior art, and FIG. 19 shows an active vibration isolator according to the first prior art and an XY stage mounted thereon. It is a figure which shows a structure.
In FIG. 18, reference numeral 201 denotes an exposure apparatus main body including a vibration isolation table, and the following sensors and actuators are attached to control the vibration isolation table. That is, AC is an acceleration sensor 251 serving as vibration detection means, PO is a position sensor 252 serving as position measurement means, SV is a servo valve that supports the apparatus main body including a heavy vibration isolation table, and PR is the pressure in the air spring. The pressure sensors 253 and LM to measure are electromagnetic motors that suppress vibrations of the vibration isolation table. Symbols representing these sensors and actuators indicate a direction and a portion of the vibration isolation table.

In FIG. 19, an XY stage 203 is mounted on the vibration isolation table 202 and supported by active mounts 204-1, 204-2, and 204-3 that incorporate sensors and actuators. The acceleration sensor AC-X1 measures vibration in the X direction in the active mount 204-1. The outputs of the acceleration sensors AC-Z1, -Z2, -Z3, -X1, -Y2, and -Y3 are led to the motion mode extraction calculation means 205. Here, the signal of the motion mode such translation and rotation of the anti-vibration table 202, a so-called motion mode acceleration signals _{(a x, a y, a} z, a θx, a θy, a θz) seeking. Next, the motion mode acceleration signal is guided to the integral compensator 206 and converted into a velocity signal for each motion mode. Subsequently, it is guided to a gain compensator 207 for adjusting damping for each motion mode, and its output signal is a motion mode drive signal (d _x , d _y , d _z , d _θx , d _θy , d _θz ). It becomes. This signal becomes a drive signal (d _z1 , d _z2 , d _z3 , d _x1 , d _y2 , d _y3 ) to be generated by each actuator of each active mount through the motion mode distribution calculation means 208, and the driver 209 with this signal Is excited. With such a configuration of the vibration control system, the vibration characteristics of the vibration isolation table 202 can be adjusted to be substantially non-interfering for each motion mode.
Further, an addition terminal is provided in the preceding stage of each driver 209 of the vibration control system for the electromagnetic motor LM, and the output of the multi-channel / non-correlated / M-sequence signal generator 210 is passed through the switch 211. Connected. The switch 211 manages on / off of connections to all the electromagnetic motors LM in each active mount. In the case of illustration, all the electromagnetic motors LM are vibrated by the M series signal.
The behavior of the vibration isolation table 202 during vibration is captured by the output signals of other sensors that measure the movement state including the vibration measurement means AC, the position measurement means PO, and the pressure measurement means PR that are mounted on the exposure apparatus main body 201. The data is collected in the data storage unit 212 together with the signal from the M-sequence signal generator 210.
Next, the time series data collected in the data storage unit 212 is filtered by the pre-filtering unit 213. Subsequently, the system identification unit 214 identifies the input / output data filtered. Here, the system identification is to derive a mathematical model using input / output data. Finally, the mathematical model acquired by the system identification unit 214 is sent to the characteristic extraction / diagnosis unit 215, where further data conversion for characteristic extraction and diagnosis is performed.

  As described above, the first prior art device identifies the characteristics of each element constituting the device by applying a pseudo irregular signal (M-sequence signal) to the system constituting the closed loop or open loop control system. In addition, self-diagnosis and abnormality detection of industrial equipment such as a semiconductor exposure apparatus can be quickly performed without sacrificing the production efficiency and operation efficiency of the apparatus.

Next, Patent Document 2 which is the second prior art will be described.
FIG. 20 is a schematic diagram illustrating a method of performing control while repeatedly correcting a transfer function performed by the positioning control method in the second conventional technique.
The second conventional technique is a positioning control method proposed for the purpose of accurately obtaining an input / output characteristic between a control command and an operation of a control target, determining the control command, and causing the control target to reliably perform a desired operation, and Device.
Specifically, as shown in FIG. 20, first, the input 1 is calculated from the desired response O using the transfer function H ₁ obtained from the known input and the response, and is taught to the positioning device as target input position data. To do. At this time, if the obtained response 1 does not converge within the desired tolerance, the frequency domain response O _i is expressed by the equation (H _{i + 1} (jω) = O _i (jω) / I _i (jω)). calculating a transfer function H ₂ used. Using this transfer function H ₂ , a new input 2 is calculated from the desired response and used as target input position data. By repeating this operation until the error converges and repeating the transfer function until a predetermined control result is obtained while correcting the transfer function, highly accurate positioning is realized. That is, the target input position data is obtained so as to be efficient and highly accurate by performing positioning control while repeatedly correcting the transfer function.

Next, FIG. 21 shows an apparatus (NC turret punch press) to which the positioning control method in the second prior art is applied and its control block diagram, and FIG. 22 shows the positioning control method in the second prior art. It shows another control block diagram of the applied device.
First, in FIG. 21, 301 is a carriage, 302 is an AC servo motor, 303 is a ball screw, 304 is a servo amplifier, and 305 is an NC control unit.
As shown in FIG. 21, the drive system of the carriage 301 as a control target is driven by an AC servo motor 302 and a ball screw 303.
Further, this positioning control is digitally fed back using an AC servo motor 302, and the AC servo motor 302 is driven and controlled by the NC control unit 305 via the servo amplifier 4.
The control process in the NC control unit 305 of the AC servo motor 302 is as follows:
First, profile data that forms step-like input position data is input, target input position command data is calculated based on the profile data, and the AC servo motor 302 is operated based on the target input position command data. At the same time, the target input position command data is fast Fourier transformed (FFT).
Next, the actual movement amount of the carriage 301 measured by the laser length measuring device is sampled, and the accuracy is determined based on an error from the target input position command data. If there is no problem in the determination result, the AC servo motor 302 is continuously driven with the target input position command data.
When the determination result exceeds the limit value, the actual movement amount data is subjected to fast Fourier transform, and a transfer function is obtained from the Fourier transform data of the target input position command data and the Fourier transform data of the actual movement amount data. Is calculated.
Next, the target input position data is calculated again from the transfer function and the Fourier transform data of the target input position command data, and the newly calculated new target input position data is subjected to inverse Fourier transform. The AC servo motor 302 is operated based on the obtained new target input position data.
Next, the actual movement amount of the carriage 301 measured again by the laser length measuring device is sampled, and the accuracy is determined based on an error from the target input position data. If there is no problem in the determination result, the AC servo motor 302 is continuously driven with the new target input position data. If the determination result exceeds the limit value, the actual movement amount data is subjected to Fourier transform, and the transfer function is again obtained from the Fourier transform data of the new target input position data and the Fourier transform data of the actual movement amount data. Is calculated again from the transfer function and the Fourier transform data of the target input position command data, and the newly calculated new target input position data is subjected to inverse Fourier transform. The AC servo motor 2 is operated based on the new target input position data obtained thereby.
Therefore, as shown in the block diagram of FIG. 21, in the conventional example, the latest target input position data is always used as the target input position data for calculating the transfer function.

In the above positioning control operation, the error between the actual movement amount data and the target input position data has converged below the limit value due to the accuracy determination of the actual movement amount data detected as shown in the block diagram of FIG. In such a case, a positioning control operation is performed on the object to be controlled with the target input position command data according to the determination by the determination means.
On the other hand, when the error between the detected actual movement amount data and the target input position data does not converge below a certain level, the first correction performed while correcting the transfer function based on the determination by the determination means is stopped, and the time data is used. The second correction for correcting the target input position data is performed, and the control target is positioned and controlled using the target input position data after the second correction as the final target input position data. The second correction is correction for calculating new target input position data based on time data of an error between the target input and the actual movement amount.

As described above, the conventional positioning control method and the apparatus add the improvement that improves the correction accuracy of the transfer function to the control method by the repetitive transfer function correction using the inverse transfer function compensation method that is a well-known method, and the target input. The position data is obtained.
JP 2000-82662 A JP 7-72933 A

  However, the system identification device and the method thereof in the first conventional technique identify characteristics of each element constituting the device by applying a pseudo irregular signal (M-sequence signal) even in a closed-loop control system. However, if a pseudo-random signal (M-sequence signal) or the like is applied to the closed loop control system, the current (torque and thrust) supplied to the actuator is saturated, so the control system is restricted and the system's original There was a problem that the response could not be grasped. In other words, there is a problem that it is impossible to grasp whether the characteristics of the machine and the control parameter setting match.

Further, the positioning control method and apparatus in the second prior art is a method in which the transfer function H _{1 first} obtained from the known input and response is corrected from the target input position data and the actual movement response data to the transfer function H _2. The transfer function is calculated, but the target input position data is determined to be efficient and highly accurate by performing positioning control while repeatedly correcting the transfer function, and the optimal control system parameters cannot be adjusted. There was a problem.

  In order to adjust the control parameters so as to be optimal for the machine, such as the positioning device in the prior art as described above. For example, it is a guideline indicating whether the machine characteristics and control parameter settings match, and it is a loop-loop frequency response characteristic that makes it possible to determine the stability of the control system by grasping the gain margin and phase margin Adoption of is desired.

  The present invention has been made in view of such a problem. By utilizing a controller model from the measured open-loop frequency response characteristics (mechanical characteristics), the loop-loop frequency response characteristics are estimated and control parameters are estimated. The control parameter sensitivity analysis device of the motor control device and the control of the motor control device capable of predicting the gain margin and the phase margin as the target frequency characteristics by the control parameter setting state. An object is to provide a parameter setting method.

In order to solve the above problem, the present invention is as follows.
The invention of the control parameter sensitivity analysis device of the electric motor control device according to claim 1 generates an instruction signal and an electric motor attached to a machine, a detecting means for detecting an operation amount of the electric motor or a detected object comprising the machine a command unit which, in the command signal and the motor control device comprising a controller for driving the electric motor by receiving the detected operating amount by the detection means, characteristic of the controller from the command signal and the detected operation amount An open loop frequency response characteristic measuring means for measuring an open loop frequency response characteristic not including
And a sensitivity analysis device that performs sensitivity analysis on the relationship between the changes in the open loop frequency response characteristics calculated by the calculation means of the motor control device.

The invention of the control parameter sensitivity analysis device for the motor control device according to claim 2 generates a command signal and a motor attached to the machine, a detecting means for detecting an operation amount of the motor or a detected object comprising the machine. machine model and command unit, in the command signal and the detection means motor control device comprising a controller for driving the electric motor by receiving the operation amount detected by, was calculated by the numerical calculation of the mass and rigidity of advance the machine to A transfer function numerical model of mechanical characteristics obtained by obtaining a transfer function between the position of the motor and the position of the detection means attached to the machine, and a controller model of the controller of the motor control device When an arithmetic means for calculating a round open-loop frequency response from said transfer function numerical model and the controller model, the control parameters of the controller It is characterized by having a relationship of changes in the round open-loop frequency response characteristic calculated by data and the arithmetic means and a sensitivity analysis device for sensitivity analysis.

According to a third aspect of the present invention, there is provided the control parameter sensitivity analyzing apparatus for the motor control apparatus according to the first or second aspect, wherein the sensitivity analyzing apparatus calculates a single loop frequency response characteristic calculated by the calculation means and a target loop. and differences in loop frequency response characteristics, more partial differential of the control parameter of the round open-loop frequency response characteristics calculated by the calculating means, estimating the amount of change in the control parameter approaches the round open-loop frequency response characteristics to the target To do.

  According to a fourth aspect of the present invention, there is provided the control parameter sensitivity analysis apparatus for the motor control apparatus according to the first or second aspect, wherein the sensitivity analysis apparatus changes the control parameter within a predetermined range and step size. In this case, a single loop frequency response characteristic is calculated, and a change in the single loop frequency response characteristic due to a change in the control parameter is estimated.

According to a fifth aspect of the present invention, in the control parameter sensitivity analyzing apparatus for an electric motor control apparatus according to the first aspect, the detecting means detects any one of the position, speed, and acceleration of the electric motor or the machine. The amount of movement.

  According to a sixth aspect of the present invention, in the control parameter sensitivity analyzing apparatus for an electric motor control apparatus according to the first or second aspect, the controller includes a plurality of the control parameters.

According to a seventh aspect of the present invention, in the control parameter sensitivity analyzing apparatus for an electric motor control apparatus according to the first or second aspect, the controller includes at least one closed loop that feeds back an operation amount detected by the detecting means. The unit of the operation amount is adjusted to the unit of the command signal, and the electric motor is controlled so that the control command and the operation amount corresponding to the control command coincide with each other.

The invention according to claim 8 is the control parameter sensitivity analysis device for the motor control device according to claim 1, wherein the arithmetic means converts the measured open loop frequency response characteristic into a transfer function numerical model. Bei example numerical modeling means is for calculating said controller model of the controller, a round open-loop frequency response from the transfer function numerical model of the open loop frequency response of the pre-Symbol motor control device.

Further, the invention of claim 9 is the control parameter sensitivity analysis apparatus of the motor control device according to claim 8, wherein the numerical modeling means includes a mechanical resonance frequency, attenuation, spring multiplier, It is digitized into a transfer function numerical model composed of parameters classified into mechanical characteristics such as load or load inertia moment.

The invention of claim 10, the control parameter sensitivity analysis apparatus smell of a motor control apparatus according to claim 2 Te, by using a transfer function numerical model of the mechanical properties and the controller model, time the response to command input An operation simulator of an electric motor control device that outputs as series data is provided.

The invention of claim 11 is the control parameter sensitivity analysis apparatus for an electric motor control device according to any one of claims 8 and 9, wherein the open loop frequency response is obtained by the controller model and the numerical modeling means. An operation simulator of an electric motor control device that outputs a response to a command input as time series data using the transfer function numerical model whose characteristics are digitized is provided.

The invention of claim 12 is the control parameter sensitivity analysis for the motor control device according to any one of claims 2, 8, 9, wherein the sensitivity analysis device, round the open-loop frequency response to targets and the difference between the round open-loop frequency response calculated by the characteristic and the operation means, the partial differential of the machine parameters of the transfer function numerical model of the control parameter and the mechanical properties of round open-loop frequency response characteristics computed by the arithmetic means Thus, the amount of change in the control parameter and the machine parameter approaching the target loop-opening loop frequency response characteristic is estimated.

The invention according to claim 13 is the control parameter sensitivity analysis apparatus for an electric motor control device according to any one of claims 1, 2, and 7, wherein the controller feeds back an operation amount detected by the detection means. and at least one closed loop to, in which a switch for opening and closing at least one or more of the closed loop.

The invention according to claim 14 is the control parameter sensitivity analysis apparatus for the motor control device according to any one of claims 1, 2, and 7, wherein the controller or the controller model has various control laws. A control law database is provided.

The invention of claim 15, an electric motor mounted on the machine, a detector for detecting the amount of operation of the detected part made of the electric motor or the machine, a command for generating a command signal, the command signal and In a method for setting a control parameter of an electric motor control device including a controller that drives the electric motor in response to the operation amount detected by the detection means, characteristics of the controller are not included from the command signal and the detected operation amount. An open loop frequency response characteristic is measured and processed , a loop open loop frequency response characteristic is calculated from the open loop frequency response characteristic and the controller model of the controller, and is calculated by the control parameter of the controller and the calculation process. In addition, a sensitivity analysis process is performed on the relationship between the changes in the open loop frequency response characteristics, and the control parameters of the controller are determined.

The invention of claim 16 includes an electric motor attached to a machine, a detecting means for detecting an operation amount of a detected object comprising the electric motor or the machine, a command device for generating a command signal, In a control parameter setting method of an electric motor control device that includes the controller that drives the electric motor in response to the command signal and the operation amount detected by the detection means, a machine model that is previously calculated by numerical calculation from the mass and rigidity of the machine Using a numerical model of a transfer function of a mechanical characteristic obtained by obtaining a transfer function between the position of the electric motor and the position of the detection means attached to the machine, and a controller model of the controller A procedure of calculating a loop frequency response characteristic, performing a sensitivity analysis process on a relationship between a control parameter of the controller and a change in a loop frequency response characteristic calculated by the calculation process, and determining a control parameter of the controller I took it.

The invention according to claim 17 is the motor control device control parameter setting method according to any one of claims 15 and 16 , wherein the sensitivity analysis process includes a target open loop frequency response characteristic and the calculation process. and the difference between the calculated round open-loop frequency response in, more and partial differential before Symbol control parameters round the open-loop frequency response characteristic calculated by the arithmetic processing approaches to round the open-loop frequency response characteristics to the target It said than is the amount of change control parameters taking estimated Then say procedures.

Further, the invention of claim 18 is the motor control device control parameter setting method according to any one of claims 15 and 16, wherein in the sensitivity analysis process, the range and step size arbitrarily determined in advance are used. The procedure is to calculate a single loop frequency response characteristic when the control parameter is changed and to estimate a change in the single loop frequency response characteristic due to the change in the control parameter.

The invention according to claim 19 controls the motor so as to optimally adjust the control parameter by using the control parameter sensitivity analysis device for the motor control device according to any one of claims 1 to 14. It features an electric motor control device.

  According to the first aspect of the present invention, even when the control parameter is set to a high gain, the open loop frequency response characteristic can be measured without causing torque saturation or thrust saturation, and the control of the controller Since the relationship between the parameter and the change in the loop response frequency characteristic is known, the target control parameter value is clarified, and the adjustment of the motor control device can be easily performed.

According to the second aspect of the present invention, if the mechanical characteristics of the controlled object of the motor control device are given as a transfer function numerical model, the stability of the control system can be determined based on the open loop frequency response characteristics.

According to the third aspect of the present invention, the amount of change in the control parameter to approximate the target one-round loop frequency response characteristic is estimated by the control parameter sensitivity analysis with respect to the target one-round loop frequency response characteristic. Therefore, the change amount to the control parameter, that is, the target control parameter value is clarified, and the adjustment of the motor control device can be easily performed.

  According to the fourth aspect of the present invention, even when the control parameter is set to a high gain, the open loop frequency response characteristic can be measured without causing torque saturation (or thrust saturation). The loop frequency response characteristics can be estimated, and the control parameter sensitivity analysis that shows the change in the open loop frequency response characteristic with respect to the change in the control parameter can predict the change in the open loop frequency response characteristic, and the motor controller can be adjusted. Easy to implement.

  According to the invention described in claim 5, it is possible to estimate the open loop frequency response characteristic regardless of the type of the detection means of the electric motor control device and the installation location.

  According to the sixth aspect of the present invention, the controller can have a plurality of control parameters.

  According to the seventh aspect of the present invention, the controller can form a plurality of closed loops.

  According to the eighth aspect of the present invention, the open loop frequency response characteristic portion can also be digitized to estimate the open loop frequency response characteristic.

  According to the ninth aspect of the invention, since the open loop frequency response characteristic is digitized into a numerical model composed of mechanical characteristic parameters, there is an effect that a factor of the mechanical characteristic can be clearly modeled.

According to the invention described in claims 10 and 11, in addition to the estimation result of the open loop frequency response characteristic, the adjustment of the motor control device can be confirmed by the operation simulation.

  According to the invention described in claim 12, the amount of change in the machine parameter can be taken into account as a factor for approaching the target loop open loop frequency response characteristic or a factor for changing the loop open loop frequency response characteristic.

Further, according to the invention described in claim 13 , since an open loop and a closed loop can be selected, normal operation, measuring a frequency response characteristic equivalent to the open loop frequency response characteristic without causing a position shift, You can choose to measure frequency response characteristics.

According to the fourteenth aspect of the present invention, it is possible to estimate the open loop frequency response characteristic when a control law different from the current state is used.

According to the fifteenth and sixteenth aspects of the present invention, the control parameters can be set by predicting the open loop frequency response characteristics.

According to the inventions of claims 17 and 18 , the amount of change in the control parameter that approximates the target one-open loop frequency response characteristic is estimated, or how much one loop-open loop frequency with respect to the amount of change in the control parameter. Estimate whether the response characteristics change.
Either sensitivity analysis process can be performed.
Furthermore, according to the invention of claim 19 , the motor is controlled so as to optimally adjust the control parameter by using the control parameter sensitivity analysis device of the motor control device according to any one of claims 1 to 14. By controlling the electric motor with the electric motor control device characterized in that, the same effects as the effects described in the first to fourteenth aspects can be obtained.

  Embodiments of the present invention will be specifically described below with reference to the drawings.

  FIG. 1 is an overall configuration diagram of a control parameter sensitivity analysis apparatus for an electric motor control apparatus according to a first embodiment of the present invention, and FIG. 2 is a control block diagram of the control parameter sensitivity analysis apparatus for the electric motor control apparatus according to the first embodiment. In this example, first, the configuration of the control parameter sensitivity analysis device of the motor control device will be described.

  In the figure, 1 is an electric motor, 2 is a detection means, 3 is a controller, 4 is a command device, 5 is a machine, 6 is an open-loop frequency response characteristic measurement means, 7 is a calculation means, 8 is an input device, and 9 is an output device. 10 is a storage device, 11 is a gain balance equation, 12 is a control parameter sensitivity analyzer, 14a and 14b are switches, 15 is a control law database, 16a and 16b are adders, 17 is an operation simulator, 23 is a controller model, 35 is a driving force detecting means, 37 is a closed-loop disturbance driving force response frequency characteristic measuring means, and 38 is a loop-opening loop frequency response characteristic calculating means. In the present embodiment, the electric motor is an example of a rotary motor, and the detection means 2 is an encoder added to the rotary motor.

The features of the present invention are as follows.
That is, the motor 1 attached to the machine 5, the detection means 2 for detecting the operation amount of the detected object comprising the motor 1 or the machine 5, the command device 4 for generating the command signal, and the command signal In an electric motor control device comprising a controller 3 for driving the electric motor 1, an open loop frequency response characteristic measuring means 6 for measuring an open loop frequency response characteristic not including the characteristic of the controller 3, and a controller of the electric motor control device The controller model 23, the calculation means 7 for calculating the open loop frequency response characteristic from the measured open loop frequency response characteristic and the controller model, the control parameter of the controller and the change of the open loop frequency response characteristic And a sensitivity analysis device 12 for analyzing the sensitivity of the relationship.

Further, the sensitivity analysis device 12 sets the target based on the difference between the one-round loop frequency response characteristic and the target one-round loop frequency response characteristic and the partial differentiation of the control parameter of the mathematical model of the single loop-frequency response characteristic. The amount of change in the control parameter that approaches the loop-opening loop frequency response characteristic is estimated.
Further, the sensitivity analyzer 12 calculates the loop-opening loop frequency response characteristic when the control parameter is changed within a predetermined range and step size, and the loop-opening loop frequency response due to the change of the control parameter. The change in characteristics is estimated.
Further, the sensitivity analysis device 12 is provided with a gain balance type that maintains a balance while changing the responsiveness of the control characteristics when having a plurality of control parameters.
In addition, the sensitivity analysis device 12 includes a machine parameter as a control parameter, a difference between a target one-round loop frequency response characteristic and the estimated one-round loop frequency response characteristic, and a formula of the estimated one-round loop frequency response characteristic. Based on the partial differentiation of the control parameters of the model, the amount of change in the control parameters including the machine parameters approaching the target loop-opening loop frequency response characteristics is estimated.

The controller 3 includes at least one closed loop, and controls the electric motor 1 so that the control command and the operation amount of each component coincide with each other in accordance with the unit system.
Further, before and after the controller 3, a driving force detection means 35 that detects a driving force that is output from the controller 3 and that becomes a torque or thrust before the adder 16b provided in at least one closed loop, and at least Closed-loop disturbance driving force response frequency characteristic for obtaining a closed-loop disturbance driving force response frequency characteristic obtained from the signal of the command generator added to the adder 16a provided in one closed loop and the signal response detected by the driving force detection means 35 The measuring means 37 is provided.
The computing means 7 includes numerical modeling means for numerically modeling the measured open loop frequency response characteristics, numerically modeling the open loop frequency response characteristics, and the controller model 23 of the controller 3 of the motor control device. Then, the open loop frequency response characteristic is calculated from the numerical model of the open loop frequency response characteristic.

Next, the procedure of the control parameter setting method will be described based on FIG. 2 and FIG. FIG. 3 is a flowchart showing the procedure of the control parameter setting method of the motor control device in the first embodiment.
First, the procedure for measuring the open loop frequency response characteristic in step ST01 will be described.
Step ST01 includes ST01a, ST01b, and ST01c. Among them, the process of acquiring the open loop frequency response at step ST01a, have become earthenware pots by measuring the open loop frequency response characteristic of the motor control device to open looped not include characteristics of the controller. In step ST01b, the motor control device is closed and the signal of the command generator is input to an adder provided in the closed loop, so that the open loop frequency response characteristic measuring means Equivalent frequency characteristics are measured. Further, in step ST01c, a transfer function numerical model of a mechanical characteristic between a drive unit driven by the electric motor of the machine equivalent to an open-loop frequency response characteristic and an operation amount detection unit attached to the detection unit of the machine is calculated. Is calculated. In this embodiment, step ST01a will be described first, step ST01b will be described in the fourth embodiment, and step ST01c will be described in the fifth embodiment.
Step ST01a will be specifically described.
In a closed loop having a feedback loop that operates the controller of the motor control device, when a command signal for obtaining a frequency response characteristic is input, torque saturation occurs, so that the switch 14 is switched to an open loop. The switches 14a and 14b give the command signal C of the command device 4 to the motor 1 and the machine 5 as a load, and the response R is detected by the detection means 2 . With the switch 14c, the feedback loop is cut to open.
The open loop frequency response characteristic measuring means 6 receives the command signal C and the response R, and obtains an open loop frequency response characteristic.

Next, a procedure for estimating the open loop frequency response characteristic from the measurement result of the open loop frequency response characteristic in step ST02 will be described.
Step ST02 includes ST02a and ST02b.
In the case of step ST02a, since the open loop frequency response characteristic H obtained by the open loop frequency response characteristic measuring means 6 is provided as digital data, the circuit is opened by the controller model 23 of the controller 3 possessed by the arithmetic unit 7. The loop frequency response characteristic Zo can be estimated.
FIG. 4 is a diagram showing the relationship among the open loop frequency response characteristic H, the controller model G, the command signal C, and the response signal R.
FIG. 5 is a diagram showing the open loop frequency response characteristic Zo. The one-round open loop frequency response characteristic Zo is a characteristic that the command signal C reaches the response signal R through the controller model G and the open loop frequency response characteristic H.
If there is a control parameter G that is a characteristic of the controller model 23, the open loop frequency response characteristic Zo is expressed by the following equation (1). If the arithmetic unit 7 includes the controller model 23 as a frequency domain model, the open loop frequency response characteristic Zo can be estimated.

  On the other hand, in the case of step ST02b, the open loop frequency response characteristic H is numerically modeled. For example, it is quantified as a well-known formula (2).

Here, s: Laplace operator, a _i , b _i : coefficients.
In the case of step ST02b, if the control model is not a frequency domain model, it is possible to use a control model equivalent to the actual controller 3 if it is digitized as in the above equation (2). For this reason, even when it is complicated to reflect in Formula (1) of the frequency domain, it is possible to estimate the open loop frequency response characteristic Zo. One-round loop frequency response characteristic Zo can be estimated from the configuration and operation of an operation simulation described later.
The one-round loop frequency response characteristic Zo can be estimated by the same procedure as the measurement of the one-round loop frequency response characteristic Zo described later.

In addition, although numerical modeling of the open loop frequency response characteristic H was demonstrated by Formula (2), you may digitize with another model.
An example of numerical modeling in step ST02b will be additionally described.
Here, FIG. 6 is a graph showing an example of the measurement result and numerical model of the open loop frequency response characteristic that can be approximated to the two-inertia system of the embodiment of the present invention. FIG. 7 is a graph showing an example of a measurement result of an open loop frequency response characteristic that can be approximated to the multi-inertia system of an embodiment of the present invention and a numerical model. FIG. 8 is a two-inertia spring / mass model diagram of the embodiment of the present invention, and FIG. 9 is a multi-inertia spring / mass model diagram of the embodiment of the present invention.
Here, the solid line indicates the measured open loop frequency response characteristic, and the broken line indicates the numerically modeled open loop frequency response characteristic.
If the open-loop frequency response characteristic can be approximated to a two-inertia system as shown in FIG. 6, the open-loop frequency response characteristic H ₁₁ obtained by driving the motor 1 in FIG. Numerical modeling can be performed using the following equation (3).

Here, J ₁ is the load inertia of the motor side, J ₂ is load inertia of the load, k is the spring constant, d is the attenuation. J is the sum of _{J 1} and _{J 2.}

Further, the expression (3) is quantified using physical quantities such as load or load inertia moments J ₁ , J ₂ , spring constant k, and damping d, but the resonance frequency ω and anti-resonance frequency obtained from the open loop frequency response characteristics It may be quantified using the characteristic quantities ω _a and the damping ratios ζ and ζ _a .
Expression (4) is a modification of Expression (3).

here,
J = J1 + J2
ω = √ (k (1 / J ₁ + 1 / J ₂ ))
ζ = (d / 2) √ ((1 / J ₁ + 1 / J ₂ ) / k)
ωa = √ (k / J2)
ζa = (d / 2) √ (J ₂ / k)
Furthermore, if the open loop frequency response characteristic can be approximated to a multi-inertia system as shown in FIG. 7, it can be numerically modeled using Equation (5).

Here, N represents the number of resonances.
In this example, the case of multi-inertia as shown in FIG. 9, the load or moment of inertia _{J 1, J 2, ····,} J N, the spring constant _{k 1,} ‥ _{k N,} attenuation _{d 1,} that ‥ _{d N} Modeling may be performed using physical quantities. However, since it is difficult to estimate and identify the multi-inertia, the resonance frequency ω _r obtained from the open-loop frequency response characteristic, the damping ratio ζ _r , the amplitude and the phase A _{Using r} , the sum of the vibrations of the one-degree-of-freedom system is quantified as equation (5) as proportional viscosity damping.
Considering the influence B of higher-order resonance, the equation (6) including the excess stiffness B may be used.

  When a general viscous damping system is applied, equation (6) becomes equation (7)

Here, U _r , V _r : amplitude and phase, and σ _r : attenuation.

Next, the sensitivity analysis in step ST03 will be described.
In the case of ST03a, which is a sensitivity analysis of a single loop frequency response characteristic with respect to a change in control parameter, a plurality of loop loop frequency response characteristics are estimated. The control parameter G is changed within a predetermined range within a predetermined range, a new open loop frequency response characteristic Zo is estimated, and graphs are superimposed and displayed on the output device 9. Thereby, the difference in the relative open loop frequency response characteristic Zo due to the difference in the control parameter G is visually clarified.
Note that when there are a plurality of control parameters G such as a multiple loop, the control parameter G may be changed according to the gain balance equation 11 that maintains the balance of the control parameters.

In the case of step ST03b, which is sensitivity analysis of a control parameter with respect to a change in the loop-opening loop frequency response characteristic, first, a target loop-opening loop frequency response characteristic Zo ′ is input from the input device 8.
In order to make the current estimated loop-opening loop frequency response characteristic Zo close to the target loop-opening loop frequency response characteristic Zo ′, the difference ΔZo according to the equation (8) is ΔZo = Zo−Zo ′ Equation (8)
Then, the partial differential of the control parameter G of the currently estimated loop-opening loop frequency response characteristic Zo consisting of the equation (1) and the control parameter change amount ΔG have the relationship of the equation (9).

  If formula (9) is written in detail,

It becomes. Here, the control parameter G is expressed as having a plurality of elements g ₁ , g ₂ ,... _Gn .
Thus, sensitivity analysis can be performed by deriving the control parameter change amount ΔG, for example, by the Gauss-Newton method.
Further, the control parameter change amount ΔG may be derived by adding a weight W to the difference between the target one-round loop frequency response characteristic Zo ′ and the current one-round loop frequency response characteristic Zo.

  Note that, when the loop-opening loop frequency response characteristic Zo is estimated, if the process goes through step ST02b, the open-loop frequency response characteristic H in the equation (1) becomes a numerical model, and these are included in the control parameters and the amount of change ΔG may be obtained. Such use will be described separately in the third embodiment.

However, when the change amount ΔG of the control parameter is derived from the equation (9) or the equation (10), the change amount ΔG may not be close to the target single loop frequency response characteristic Zo ′. That is, there is a case where the change amount ΔG is calculated such that ΔZo increases although the sensitivity of the control parameter change to the open loop frequency response characteristic Zo ′ is high.
In this case, the control parameter sensitivity analysis result is converged by iterative calculation. The control parameter is sequentially updated as in Expression (12), and the converged change amount ΔG is derived.
_Gn = _Gn-1 + (DELTA) G Formula (12)
When ΔZo increases from the previous value, the control parameter is sequentially updated as shown in Equation (13) so that ΔZo decreases.
_Gn = _Gn-1 + e * (DELTA) G Formula (13)
Where 0 <e <1
As described above, it is possible to predict the open loop frequency response characteristic after changing the control parameter.
The control parameter change amount ΔG thus obtained can be visually confirmed by the display device.

  As described above, the sensitivity analysis of the frequency response characteristic with respect to the change of the control parameter at step ST03a or the sensitivity analysis of the control parameter with respect to the change of the frequency response characteristic at step ST03b is performed, and a guideline indicating the open loop frequency response characteristic, that is, the operation performance. The relationship of control parameters to

In step ST04, the control parameters of the controller are set and confirmed so as to bring out the target performance of the actual motor control device.
The switches 14a and 14b shown in FIGS. 1 and 2 form a closed loop so that the command signal enters the controller 3 and feeds back the response of the detection means 2, and the motor is actually operated to confirm the operation performance.

If a numerical model of the open-loop frequency response characteristic H is created in step ST02b, prior to actual machine verification in step ST04b, it may be preliminarily examined by operation simulation by numerical calculation in step ST04a.
FIG. 10 is a configuration block diagram of the operation simulation.
FIG. 10 includes a commander model 24, a controller model 23, and motor and load machine models 21 and 25, constituting a feedback loop, and inputting a command signal output from the commander model 24 to the controller model 23. The output from the controller model 23 is input to the electric motor and load machine models 21 and 25, and the output from the electric motor and load machine models 21 and 25 is input to the controller model 23 so as to be subtracted from the command signal. A so-called motor control device model is shown. Note that a detection means model for detecting the operation amount of the machine is not shown. Since the above motor control device model has a configuration simulating an actual motor control device, if a numerical model of an open loop frequency response characteristic is created in step ST02b, this numerical model is used as the motor and the load machine. If it is a model, an operation simulation can be performed.
This is the same configuration as the block diagram of FIG.
In actual FPGA ST04 or operation simulation ST04b, referring to the plurality of round open-loop frequency response estimated in step ST0 2c, to the control parameters of the controller may be arbitrarily determined, the target obtained by the step ST03b The amount of change approaching the single loop frequency response characteristic Zo ′ may be reflected in the control parameter of the controller.
In addition, as none require complicated operation of the operator, the output value of the step ST03 may be automatically be reflected in the set value used in the actual FPGA ST0 4 or operation simulation ST04 b.
Further, the result of the actual machine verification in step ST04 can also visually confirm the response of the detection means 2 by the output device 9 which is a display device.

When the actual machine verification ST04 is completed, in order to quantitatively check the stability of the control system, it is possible to measure the open loop frequency response characteristic in ST05 and grasp the gain margin / phase margin.
FIG. 11 is a simplified block diagram illustrating the measurement of the open loop frequency response characteristics.
FIG. 11 is an excerpt of the necessary parts from FIGS. 2, 4, and 9. Here, C is a command signal as in FIG. 4, R is a response signal, H is an open loop frequency response characteristic, and G is a controller model. b is compensation applied to the feedback system of the controller model, and includes unit conversion and the like. D is a disturbance input and T is a torque. Figure 11 is a configuration equivalent to the simulation of Figure 10.
In order to obtain the single loop open loop frequency response characteristic, the closed loop disturbance driving force response frequency characteristic measuring unit 37 shown in FIG. 1 is used to measure the closed loop disturbance driving force response frequency characteristic, and the single loop open loop frequency response characteristic calculating unit 38 is used. A loop open loop frequency response characteristic is calculated from the closed loop disturbance driving force response frequency characteristic.
The closed-loop disturbance driving force response frequency characteristic Zc can be obtained as shown in Expression (14) if the disturbance input D and the torque T in FIG.

  Further, the open loop frequency response characteristic Zo and the closed loop disturbance driving force response frequency characteristic Zc have a relationship as shown in Expression (15).

  That is, the open loop frequency response characteristic Zo is expressed by Equation (16).

The measurement of the open loop frequency response characteristic will be described with reference to FIGS.
Since the motor control device is in a closed loop, the switch 14c in FIG. 2 is closed to form a closed loop.
The switch 14a is set to the switch 14b side in order to give the command signal C of the command device 4 as the disturbance D. The switch 14b acts to give the torque T and disturbance D of the controller 3 to the adder 16b, and gives driving force to the electric motor 1 and the machine 5. The detection means 2 detects the response of the electric motor 1 or the machine 5, and feeds it back from the adder 16a to the controller 3 via the switch 14a.
The closed-loop disturbance driving force response frequency characteristic measuring unit 37 measures the closed-loop disturbance driving force response frequency characteristic using the equation (14). The open loop frequency response characteristic calculating means 38 calculates the open loop frequency response characteristic Zo represented by the equation (16) according to the equation (15) using the closed loop disturbance driving force response frequency property Zc.
If the loop open loop frequency response characteristic Zo is graphed on the output device 9 and shown in a Bode diagram or a Nyquist diagram, the gain margin and the phase margin can be determined, so that the stability of the control system can be determined.

  Further, in the storage device 10, the open loop frequency response characteristic, a numerical model of the open loop frequency response characteristic, the estimated one open loop frequency response characteristic, the one open loop frequency response characteristic when a control parameter is changed, The control parameter of the controller model, the measured loop open loop frequency response characteristics, the control parameter of the controller, the amount of change of the control parameter estimated to approach the target, or the configuration specifications of the motor control device, etc. Thus, it is possible to perform estimation of the open loop frequency response characteristics and analysis of control parameter sensitivity of the motor control device at different dates and places.

Furthermore, since the control model database 15a of the controller model is provided, it is possible to estimate the open loop frequency response characteristic when a control rule other than the current controller control law is applied, and to control the change in the frequency response characteristic in that case. Parameter sensitivity analysis or sensitivity analysis of frequency response characteristics with respect to changes in control parameters can be performed.
The control law may be changed by the input device 8 from the control law database 15b of the controller.

  Thus, the control parameter sensitivity analysis apparatus of the motor control apparatus according to the present embodiment includes the open loop frequency response characteristic measurement means 6, the calculation means 7, the gain balance equation 11, the sensitivity analysis apparatus 12, and the control law database. 15 has a control parameter sensitivity analysis function, etc., so that the loop response frequency response characteristic is estimated, and the sensitivity analysis of the control parameter with respect to the change of the frequency response characteristic or the sensitivity analysis of the frequency response characteristic with respect to the change of the control parameter Therefore, it is possible to grasp the relationship of the control parameter to the loop response frequency characteristic, that is, the guideline indicating the operation performance, and the stability of the control system can be determined by the loop loop frequency response characteristic. It is possible to improve the operation performance by adjusting to be optimal.

In the second embodiment, an example in which the electric motor is a translational linear motor will be described.
FIG. 12 is a diagram showing the configuration of the control parameter sensitivity analysis apparatus of the motor control apparatus according to the second embodiment of the present invention.
The block diagram is the same as FIG. 2 of the first embodiment. In the figure, 1 is an electric motor, 2 is a detection means, 3 is a controller, 4 is a command device, 5 is a machine, 6 is an open-loop frequency response characteristic measurement means, 7 is a calculation means, 8 is an input device, and 9 is an output device. 10 is a storage device, 11 is a gain balance type, 12 is a sensitivity analysis device, 15 is a control law database, 16 is an adder, 17 is an operation simulator, 23 is a controller model, 37 is a closed-loop disturbance driving force response frequency characteristic measurement Means 38 is a loop-opening loop frequency response characteristic computing means. The electric motor 1 is a translational linear motor, and the detection means 2 is a linear encoder that detects the amount of movement of the translational linear motor mover and the load of the machine 5, which is different from the first embodiment.
The second embodiment differs from the prior art in that, like the first embodiment, the open loop frequency response characteristic measuring means 6, the calculating means 7, the gain balance equation 11, the sensitivity analysis device 12, and the control law database. 15, an operation simulator 17, a controller model 23, a closed-loop disturbance driving force response frequency characteristic measuring unit 37, and a one-open loop frequency response characteristic calculating unit 38.
In the second embodiment, the configuration of the electric motor 1 and the detection means 2 is different from that of the first embodiment, but the block diagram is the same as FIG. 2 of the first embodiment. For this reason, the operations from step ST01 to ST05 are the same as those in the first embodiment.
Thus, since it has composition like a 1st example, the same effect as a 1st example can be acquired. Even if the electric motor 1 is a rotary motor or a translational linear motor, the same effect can be obtained.

In the third embodiment, an example of full-closed control in which the electric motor is a rotary motor and load side detection means is added will be described. An additional explanation will be given for sensitivity analysis of control parameters, and a case will be described in which machine parameters are included in the control parameters and control parameter sensitivity analysis is performed.
FIG. 13 is a diagram showing the configuration of the control parameter sensitivity analysis apparatus of the motor control apparatus according to the third embodiment of the present invention.
FIG. 14 is a block diagram showing a motor control device portion of the control parameter sensitivity analysis device of the motor control device according to the third embodiment of the present invention.
13 and 14, 1 is an electric motor, 2 a and 2 b are detection means, 3 is a controller, 4 is a command device, 5 is a machine, 6 is an open loop frequency response characteristic measurement means, 7 is a calculation means, and 8 is an input. Device, 9 output device, 10 storage device, 11 gain balance type, 12 sensitivity analysis device, 14 switch, 15 control law database, 16 adder, 17 operation simulator, 23 controller model, Reference numeral 35 denotes a driving force detection means, 37 denotes a closed loop disturbance driving force response frequency characteristic measuring means, and 38 denotes a one-open loop frequency response characteristic calculation means.
FIG. 13 shows a block diagram of only the motor controller.
Further, the electric motor 1 is a rotary motor, and the detection means 2 is an encoder 2a added to the rotary motor and a linear encoder 2b that detects an operation amount of a load of the machine 5, and a rotation that detects a speed. This is a full-closed control system including both a type encoder 2a and a linear encoder 2b for detecting a position.
Further, the controller 3 constitutes a multiple loop system of a position loop and a speed loop. The position loop has a position loop gain Kp for the position loop, and a position proportionality having a speed loop gain Kv and an integration time constant Ti for the speed loop. -The speed proportional / integral (P-PI) control is configured, the torque filter Tf is additionally provided, and four control parameters G such as Kp, Kv, Ti, and Tf are configured.
Such a point is different from the first and second embodiments.
The present invention is different from the prior art in that, like the first and second embodiments, the open loop frequency response characteristic measuring means 6, the calculating means 7, the gain balance equation 11, the sensitivity analyzing apparatus 12, the control law This is a part provided with a database 15, an operation simulator 17, a controller model 23, a driving force detection means 35, a closed loop disturbance driving force response frequency characteristic measuring means 37, and a one-round open loop frequency response characteristic calculating means 38.

Also in this embodiment, the operations from step ST01 to ST04 are basically the same as those in the first embodiment.
The description will focus on differences from the first and second embodiments due to the configuration of the detection means 2a, 2b and the multiple loop system.
The measurement of the open-loop frequency response characteristics of the step ST01, the mechanical properties of the open-loop frequency response H _V of the mechanical properties to the motor speed-detecting means 2a from the torque, from the torque to the machine load position and detecting means 2b two open-loop frequency response characteristic different from that open-loop frequency response H _D is present.
Therefore, similarly to the first and second embodiment measures the open-loop frequency response H, to obtain two kinds of open-loop frequency response H _V and H _D.

In step ST02, one loop open loop frequency response characteristic Zo is estimated in the same manner as in the first and second embodiments, but two types of one loop loop frequency response characteristic Zo are estimated based on the result via ST02a or ST02b.
Control parameter corresponding to the formula (1) it is, Kp, Kv, Ti, so constitute a position proportional-velocity proportional integral (P-PI) control consisting of Tf, and the open-loop frequency response _{H V,} open When denoted by the loop frequency response characteristic H _D, round the open-loop frequency response characteristic Zo is as equation (17).

Here, the open-loop frequency response H _V, the open-loop frequency response H _D different unit system response.
As described in the first embodiment, since the closed-loop disturbance driving force response frequency characteristic Zc and the one-open loop frequency response characteristic Zo have the relationship of the above equation (15), the closed-loop disturbance driving force response frequency property Zc is an equation. (18)

As described above, the open loop frequency response characteristic Zo can be estimated from the open loop frequency response characteristic (H _D , H _V ) and the control parameter G (Kv, Kp, Ti, Tf) as in the first embodiment. it can.
Moreover, the open-loop frequency response H _D in step _ST02b, in terms of the H _V were quantified in the same manner as the first embodiment, it may be estimated round open-loop frequency response characteristics Zo.
When the open-loop frequency response characteristics H _D and H _V are digitized into the two-inertia model as shown in FIG. 8, when the J ₁ side is driven and the response on the J ₂ side is detected, the mechanical load position / detection means is calculated from the torque. open-loop frequency response H _D serving as mechanical properties until 2b becomes equation (19).

When driving the J ₁ side detects the response of J ₁ side is equivalent to the formula shown in the first embodiment (3). This corresponds to the open-loop frequency response H _V as the mechanical properties of the torque until the motor speed-detecting means 2a.
When ST02b is completed as described above, the open loop frequency response characteristic Zo can be estimated as in the first embodiment.
As in the first embodiment, the open loop frequency response characteristic may be converted into a numerical value as a multi-inertia model to estimate the open loop frequency response characteristic Zo.

  In step ST03, the sensitivity analysis of the frequency response characteristic with respect to the change of the control parameter in step ST03a or the sensitivity analysis of the control parameter with respect to the change of the frequency response characteristic in step ST03b is performed as in the first embodiment.

Here, an additional description will be given of an example in which sensitivity analysis including elements of the numerical model is performed when the open loop frequency response characteristics are digitized in STP02b.
Similar to the first embodiment, both step ST03a and step ST03b can be performed.
In the sensitivity analysis of the frequency response characteristic with respect to the change of the control parameter in step ST03a, as in the first embodiment, the control parameter G (Kp, Kv, Ti, Tf) can be used to estimate a new open loop frequency response characteristic Zo.
If the open loop frequency response characteristic is digitized in step STP02b, the control parameter G and the machine parameter can be changed to estimate a new one-round open loop frequency response characteristic Zo.

When the open-loop frequency response characteristic is digitized into the two-inertia model as described above, the load inertia moment J ₁ on the motor side, the load inertia moment J ₂ on the load side, the spring constant k, and the damping d are changed and new values are obtained. One loop frequency response characteristic Zo can be estimated.
As described in the first embodiment, the new inertial loop frequency response characteristic Zo is obtained by changing the characteristic values of the two-inertia model such as the resonance frequency ω, the anti-resonance frequency ω _a , and the damping ratio ζ, ζ _a. May be estimated.
When the open loop frequency response characteristic is approximated to a multi-inertia system, even if each resonance frequency ω _r , damping ratio ζ _r , amplitude and phase A _r , total load inertia moment J, or excess stiffness B is changed. If the multi-inertia system is not a proportional viscous damping system but a general viscous damping system, each resonance frequency ω, damping ratio ζ, amplitude and phase U and V, total load inertia moment J, or excess stiffness B, May be changed

In the sensitivity analysis of the control parameter with respect to the change of the frequency response characteristic in step ST03b, the target one-round loop frequency response characteristic Zo ′ is inputted and the target one-round loop frequency response characteristic Zo ′ is input as in the first embodiment. The amount of change in the control parameter G to be close to is obtained, and the sensitivity analysis can be performed by including the machine parameter in the control parameter G.
The machine parameter is obtained if the open loop frequency response characteristic is numerically modeled by STO2b.
First, the sensitivity analysis of only the control parameter G (Kp, Kv, Ti, Tf) is shown as in equation (20).

Such sensitivity analysis may be performed on the open loop frequency response characteristic Zo.
If the open-loop frequency response characteristic is a two-inertia model, load inertia moment J ₁ , load-side load inertia moment J ₂ , spring constant k, damping d, resonance frequency ω, anti-resonance frequency ω _a , and damping thereof In the case of the ratio ζ, ζ _a , and the multi-inertia model, the resonance frequency ω _r , the damping ratio ζ _r , the amplitude and phase A _r , the total load inertia moment J, or the excess stiffness B may be used. If the system is not a proportional viscous damping system but a general viscous damping system, each resonance frequency ω _r , damping ratio ζ _r , amplitude and phase U _r and V _r , total load inertia moment J, or excess stiffness B Sensitivity analysis can be included in the control parameter G (Kp, Kv, Ti, Tf).
If the open-loop frequency response characteristic is a two-inertia model, sensitivity analysis can be performed based on Equation (21).

  Note that if the two-inertia model is digitized as in the above equation (4), the sensitivity analysis can be performed based on the equation (22).

  If the open loop frequency response characteristic is approximated to a multi-inertia system and a proportional viscous damping system is used, sensitivity analysis can be performed based on Equation (23).

  Thus, in step ST03, as in the first embodiment, sensitivity analysis can be performed to extract appropriate control parameters. If the open loop frequency response characteristic H is quantified, sensitivity analysis including machine parameters can be performed, and the improvement on the machine side for achieving the target operation performance can be understood.

In step ST04, as in the first embodiment, it is possible to check actual machine operation and use an operation simulator.
In this example, the control parameters of the controller are four feedback parameters Kp, Kv, Ti, and Tf. However, feedforward parameters may be configured, or a notch filter may be added. .

  In this example, the detection means for detecting the speed and the detection means for detecting the position (displacement) are used. However, a detection means for detecting the acceleration may be used.

In the fourth embodiment, the measurement of the open loop frequency response characteristic in step ST01b will be additionally described.
FIG. 15 is a block diagram showing a motor control device portion of a control parameter sensitivity analysis device for a motor control device according to a fourth embodiment of the present invention.
1 is an electric motor, 2a and 2b are detection means, 3 is a controller, 4 is a command device, 5 is a machine, 14a, 14b and 14c are switches, 16a, 16b, 16c and 16d are adders, and 35 is a driving force detection means. It has become.
Further, the controller 3 constitutes a multiple loop system of a position loop and a speed loop. The position loop has a position loop gain Kp for the position loop, and a position proportionality having a speed loop gain Kv and an integration time constant Ti for the speed loop. -The speed proportional / integral (P-PI) control is configured, the torque filter Tf is additionally provided, and four control parameters G such as Kp, Kv, Ti, and Tf are configured.
Other configurations are the same as those in the first, second, and third embodiments. As shown in FIGS. 1, 2, 12, and 13, the output device 9, the storage device 10, and the gain balance type are used. 11, sensitivity analysis device 12, control law database 15, operation simulator 17, controller model 23, closed-loop disturbance driving force response frequency characteristic measuring means 37, one-round loop frequency response characteristic calculating means 38, and the like.
Unlike the first and second embodiments, the configuration of a multi-loop system is shown. The third embodiment is a full-closed control system, but FIG. 15 is different in that it is a semi-closed control system.
The difference between the present invention and the prior art is the same as in the first, second, and third embodiments, but only the motor control device portion is shown in FIG.

In the figure, there is a route in which the command signal C of the command device 4 is added to the position loop by the switch 14 a and a route to be added immediately before the motor 1.
In the measurement of the open loop frequency response characteristic in step ST01, it is possible to obtain a frequency response characteristic equivalent to the open loop frequency response characteristic in the closed loop state in ST01b.
The switch 14c is turned on to make a closed loop state, and the response of the detection means 2 is input (-) by the adder 16a. While there is no input to the controller 3 and the controller 3 operates so as to keep the input of “0”, the command signal C of the command device 4 is input to the adder 16b immediately before the motor 1 by the switch 14a, and the open loop The frequency response characteristic measuring means 6 receives the command signal C, the response r, and a round of torque (thrust in the case of a linear motor) T, and obtains an open loop frequency response characteristic.
A frequency response characteristic H equivalent to the open loop frequency response characteristic is obtained by the equation (24).

In this way, the frequency response characteristic H is obtained from the frequency analysis result of the sum of the command signal C and the round of torque T and the frequency analysis result of the response r.
Since step ST01b can obtain the same frequency response characteristic H as ST01a, steps ST02, ST03, ST04, and ST 0 5 can be performed in the same manner as in the first to third embodiments.

In the configuration of FIG. 15, the open loop frequency response characteristic measured in ST 0 5 is as follows.
First, as described in the first embodiment, the closed-loop disturbance driving force response frequency characteristic is expressed by Equation (25).

  From the relationship of the equation (15), the open loop frequency response characteristic Zo becomes the equation (26).

  As described above, step ST01 can measure an open loop frequency response characteristic even in a closed loop state. Further, a single-open loop frequency response characteristic Zo of the semi-closed loop system is shown.

Of the present invention, calculation of a loop-opening loop frequency response characteristic using a transfer function numerical model obtained by numerical calculation and other frequency characteristics will be described. That is, step ST01c in FIG. 3 will be described.
If one-round open-loop frequency response characteristics are calculated using a transfer function numerical model obtained by numerical computation, the open-loop frequency response characteristics are the mechanical characteristics of the controlled object. If the modeled characteristics are given, the open loop frequency response characteristics can be estimated as in the first embodiment.
In this case, since the machine model is provided, the sensitivity analysis including the control parameter and the machine parameter can be performed as described in the third embodiment.
In addition, there is a feature that an operation simulator can be used.

FIG. 16 is a result diagram of the finite element method according to the fifth embodiment of the present invention.
An electric motor driven by the electric motor control device is incorporated in the machine, and the electric motor is controlled by feeding back the response of the detection means. The characteristics of this machine are calculated by the finite element method. The machine's natural frequency and vibration mode can be obtained by putting the dimensions and physical properties of the machine into a computer, incorporating them as a machine model, and defining and solving various conditions such as boundary conditions according to the actual machine. If the natural frequency and vibration mode are obtained, the machine model can be quantified, so if the transfer function between the location of the motor and the location of the detection means is output, the mechanical properties corresponding to the measured open-loop frequency response characteristics The transfer function numerical model can be grasped.
FIG. 16 is a transfer function numerical model diagram of the fifth embodiment of the present invention. FIG. 17 shows a Bode diagram of the transfer function numerical model.
The transfer function numerical model can be obtained by inputting parameters output as a result of the finite element method, for example, into the equation (6).
If this transfer function numerical model is input to the arithmetic unit of the control parameter sensitivity analysis device of the motor control device shown in FIG. 1, for example, the open loop frequency response characteristic can be estimated using the control model.
The present invention is different from the prior art in that a transfer function numerical model, which is a mechanical characteristic, is calculated not by an actual machine but by a numerical operation by a finite element method. This also applies to portions different from the first to fourth embodiments.

  Note that the target machine may be a drawing level without an actual machine, or may be an existing machine. The obtained transfer function numerical model is regarded as the open-loop frequency response characteristic, and the one-round open-loop frequency response characteristic can be estimated in the same manner as in the first to third embodiments.

  As described above, since the open loop frequency response characteristic can be estimated from the transfer function numerical model, the control parameter can be estimated without an actual machine, and the machine performance can be predicted. Further, since the sensitivity analysis of the stability of the control including the machine parameters can be performed, the improvement including the machine can be examined, and the measures for improving the performance can be implemented by combining the mechanical system and the control system.

The flowchart of FIG. 3 is also a flowchart showing a processing procedure for setting the control parameter of the motor control device in the control parameter sensitivity analysis device of the motor control device. The method of the present invention will be described step by step with reference to this figure.
First, in step ST01, the control target (mechanical characteristics) is grasped. Next, in step ST02, a loop open loop frequency response characteristic is estimated. Further, in step ST03, sensitivity analysis of the relationship between the control parameter and the open loop frequency response characteristic is performed. In step ST04, the actual operation of the motor control device is confirmed.
Finally, in step ST05, the loop response frequency response characteristic is measured to confirm the control system stability.
In this way, it is easy to select the control parameter because the control target (mechanical characteristics) is grasped and the tendency of the result due to the control parameter change is known in advance in consideration of the influence degree of the mechanical system. Furthermore, since the loop response frequency response characteristic is finally measured, the gain margin and the phase margin can be observed, and the appropriateness of the control parameters can be confirmed, so that the optimal control parameters of the motor control device can be set.

  After measuring the open-loop frequency response characteristics, by estimating the open-loop frequency response characteristics by numerical calculation, the load on the machine that occurs when measuring the frequency-response characteristics is minimized, and the open-loop frequency response characteristics are reduced. It can also be applied to the use of adjusting the control parameters of the controller while checking.

1 is an overall configuration diagram of a control parameter sensitivity analysis apparatus for an electric motor control apparatus according to a first embodiment of the present invention. Control block diagram of control parameter sensitivity analysis device of electric motor control device in first embodiment The flowchart figure which shows the procedure of the control parameter setting method of the motor control apparatus in 1st Example. The figure which shows the relationship between the open loop frequency response characteristic H, the controller model G, the command signal C, and the response signal R Diagram showing frequency response characteristics of open loop The graph which shows one example of the measurement result of the open loop frequency response characteristic which can be approximated to 2 inertial systems of the Example of this invention, and a numerical model The graph which shows one example of the measurement result of the open loop frequency response characteristic which can be approximated to the multi inertia system of the Example of this invention, and a numerical model 2 Inertia system spring, mass, and model diagram of an embodiment of the present invention FIG. 2 is a diagram showing a spring, mass, and model of a multi-inertia system according to an embodiment of the present invention Configuration block diagram of operation simulation of embodiment of the present invention A simplified block diagram illustrating a method for measuring a loop open loop frequency response characteristic of an embodiment of the present invention The figure which shows the structure of the motor control device control parameter sensitivity analysis apparatus of 2nd Example of this invention. The figure which shows the structure of the motor control device control parameter sensitivity analysis apparatus of 3rd Example of this invention. The block diagram which shows the motor controller part of the motor controller control parameter sensitivity analysis apparatus of 3rd Example of this invention. The block diagram which shows the motor controller part of the motor controller control parameter sensitivity analyzer of 4th Example of this invention Result diagram of the finite element method according to the fifth embodiment of the present invention Transfer function numerical model diagram of the fifth embodiment of the present invention The block diagram which showed the system identification apparatus and method applied to the active vibration isolator in the semiconductor manufacturing apparatus of 1st prior art, The figure which shows the structure of the active vibration isolator in 1st prior art, and the XY stage mounted in it Schematic diagram illustrating a method for performing control while repeatedly correcting a transfer function performed by the positioning control method in the second conventional technique NC turret punch press applying the positioning control method in the second prior art and its control block diagram NC turret punch press to which the positioning control method in the second prior art is applied and other control block diagrams

DESCRIPTION OF SYMBOLS 1 Electric motor 2 2a, 2b Detection means 3 Controller 4 Commander 5 Machine 6 Open loop frequency response characteristic measurement means 7 Calculation means 8 Input device 9 Output device 10 Storage device 11 Gain balance type 12 Sensitivity analysis device 13 Commander controller 14 , 14a, 14b Switches 15, 15a, 15b Control law database 16 Adder 17 Operation simulator 21 Electric motor model 22 Detection means model (not shown)
23 Controller model 24 Commander model 25 Machine model 31 Position loop gain 32 Speed loop gain 33 Integral time constant 34 Torque (thrust) filter 35 Driving force detecting means 36 Closed loop disturbance frequency response characteristic measuring means 37 Closed loop disturbance driving force response frequency characteristic Measuring means 38 One-round open loop frequency response characteristic calculating means

Claims (19)

An electric motor attached to the machine;
Detecting means for detecting an operation amount of a detection object comprising the electric motor or the machine;
A command device for generating a command signal;
In the motor control device comprising the command signal and a controller that drives the motor in response to the operation amount detected by the detection means ,
An open loop frequency response characteristic measuring means for measuring an open loop frequency response characteristic not including the characteristics of the controller from the command signal and the detected operation amount ;
A controller model of the controller of the motor control device;
Calculating means for calculating a round open-loop frequency response from said open loop frequency response characteristics measured with the controller model,
A control parameter sensitivity analysis apparatus for an electric motor control apparatus, comprising: a sensitivity analysis apparatus that performs sensitivity analysis on a relationship between a control parameter of the controller and a change in a single loop frequency response characteristic calculated by the calculation means . An electric motor attached to the machine;
Detecting means for detecting an operation amount of a detection object comprising the electric motor or the machine;
A command device for generating a command signal;
In the motor control device comprising the command signal and a controller that drives the motor in response to the operation amount detected by the detection means ,
A mechanical function transfer function obtained by obtaining a transfer function between the position of the electric motor and the position of the detection means attached to the machine, using a machine model calculated in advance by numerical calculation from the mass and rigidity of the machine. A numerical model;
A controller model of the controller of the motor control device;
Calculating means for calculating a round open-loop frequency response characteristic from said controller model and the transfer function numerical model,
A control parameter sensitivity analysis apparatus for an electric motor control apparatus, comprising: a sensitivity analysis apparatus that performs sensitivity analysis on a relationship between a control parameter of the controller and a change in a single loop frequency response characteristic calculated by the calculation means . The sensitivity analysis apparatus, a difference in the round open-loop frequency response characteristics to cycle the open loop frequency response characteristic and the target calculated by the calculation means, of the control parameters of the round open-loop frequency response characteristics computed by the arithmetic means by a partial differentiation, control parameters sensitivity analysis for the motor control device according to claim 1 or 2, characterized in that to estimate the amount of change in the control parameter approaches the round open-loop frequency response characteristics as a target. The sensitivity analysis apparatus calculates a loop-opening loop frequency response characteristic when the control parameter is changed within a predetermined range and step size, and the loop-opening loop frequency response characteristic due to the change in the control parameter The control parameter sensitivity analysis apparatus for an electric motor control apparatus according to claim 1, wherein a change in the control parameter is estimated. 2. The control parameter sensitivity analysis apparatus for an electric motor control device according to claim 1, wherein the detection unit detects any one of a position, a speed, and an acceleration of the electric motor or the machine and sets the operation amount. 3.   The control parameter sensitivity analysis apparatus for an electric motor control device according to claim 1, wherein the controller includes a plurality of the control parameters. The controller includes at least one closed loop that feeds back an operation amount detected by the detection unit, and adjusts the unit of the operation amount to a unit of the command signal so that the control command and the operation amount corresponding to the control command coincide with each other. The control parameter sensitivity analysis apparatus for the motor control device according to claim 1, wherein the motor is controlled as described above. It said calculating means, e Bei numerical modeling means for converting the open loop frequency response characteristics measured transfer function numerical model, the controller model and the open-loop frequency response characteristics of the controller before Symbol motor control device The control parameter sensitivity analysis apparatus for an electric motor controller according to claim 1, wherein a loop-opening loop frequency response characteristic is calculated from the transfer function numerical model . The numerical modeling means quantifies the open loop frequency response characteristic into a transfer function numerical model composed of parameters classified into mechanical characteristics such as mechanical resonance frequency, damping, spring multiplier, load or load inertia moment. The control parameter sensitivity analysis apparatus for an electric motor control apparatus according to claim 8, wherein: The motor control device according to claim 2 , further comprising an operation simulator of the motor control device that outputs a response to a command input as time-series data using the controller model and a transfer function numerical model of the mechanical characteristics. Control parameter sensitivity analyzer. An operation simulator of an electric motor control device that outputs a response to a command input as time series data using the controller model and the transfer function numerical model in which the open loop frequency response characteristic is converted into a numerical value by the numerical modeling means. 10. The control parameter sensitivity analysis device for an electric motor control device according to claim 8, further comprising: The sensitivity analysis apparatus, round round the difference between the open loop frequency response calculated by the open-loop frequency response of the operational unit, the control parameter of the round open-loop frequency response characteristics computed by the computing means for the targets and by the partial differentiation of the machine parameters of the transfer function numerical model of the mechanical properties, according to claim 2, characterized in that to estimate the amount of change in the control parameter and the machine parameters of round closer to the open-loop frequency response characteristics to the target 10. The control parameter sensitivity analysis device for an electric motor control device according to any one of claims 8 and 9 . The said controller is provided with the switch which opens and closes at least 1 or more closed loop which feeds back the operation amount detected by the said detection means, and at least 1 or more of said closed loop. The control parameter sensitivity analysis apparatus of the motor control apparatus of any one of Claims 1.   8. The control parameter sensitivity analysis apparatus for an electric motor control device according to claim 1, wherein the controller or the controller model includes a control law database in which various control laws are arranged. . An electric motor attached to the machine;
Detecting means for detecting an operation amount of a detection object comprising the electric motor or the machine;
A command device for generating a command signal;
In a control parameter setting method for an electric motor control device comprising a controller that drives the electric motor in response to the command signal and an operation amount detected by the detection means ,
An open loop frequency response characteristic that does not include the characteristics of the controller from the command signal and the detected operation amount is measured,
And a controller model of the controller and the open loop frequency response characteristics, a round open-loop frequency response characteristics and processing,
A sensitivity analysis process is performed on a relationship between a control parameter of the controller and a change in a single loop frequency response characteristic calculated by the calculation process,
A control parameter setting method for an electric motor control device, wherein the control parameter is processed by a procedure of determining a control parameter of the controller. An electric motor attached to the machine;
Detecting means for detecting an operation amount of a detection object comprising the electric motor or the machine;
A command device for generating a command signal;
In a control parameter setting method for an electric motor control device comprising a controller that drives the electric motor in response to the command signal and an operation amount detected by the detection means,
A mechanical function transfer function obtained by obtaining a transfer function between the position of the electric motor and the position of the detection means attached to the machine, using a machine model calculated in advance by numerical calculation from the mass and rigidity of the machine. A numerical model;
Compute a loop open loop frequency response characteristic from the controller model of the controller,
A sensitivity analysis process is performed on a relationship between a control parameter of the controller and a change in a single loop frequency response characteristic calculated by the calculation process,
A control parameter setting method for an electric motor control device, wherein the control parameter is processed by a procedure of determining a control parameter of the controller. The sensitivity analysis process, pre SL control parameters round the open-loop frequency response characteristic and the difference between the round open loop frequency response characteristic calculated by the arithmetic processing of round the open loop frequency response calculated by the calculation process of the target control parameter setting method of the motor control device according to claim 15 or 16, characterized in that for estimating the higher amount of change the control parameter round closer to the open-loop frequency response characteristics as a target to the partial differential of. In the sensitivity analysis process, a one-round loop frequency response characteristic when the control parameter is changed within a predetermined range and step size is calculated, and the one-loop frequency response characteristic due to the change in the control parameter is calculated. The control parameter setting method of the motor control device according to claim 15, wherein the change is estimated. By using the control parameter sensitivity analysis for the motor control device according to any one of claims 1 to 14, a motor control device and controls the electric motor to optimally perform control parameter adjustment .