JP2022077267A - Mechanical learning device and motor control device connected to the same - Google Patents

Abstract

To provide a mechanical learning device and a motor control device, which achieve highly accurate speed control and current control by executing a setting value correction and a control gain correction in a vector control system of a permanent magnet synchronous motor.SOLUTION: A mechanical learning device 101 in the present disclosure is connected to a motor control device 100 controlling a permanent magnet synchronous motor 1 to perform machine learning of the motor control device 100. In a motor shaft of the permanent magnet synchronous motor 1, there are provided a speed response data generation unit 20 that generates speed response teacher data and speed response test data on the basis of an inertia numeric constant and an inertia correction value, previously set, and a speed response classification unit 21 performing a class classification of speed responsiveness of the permanent magnet synchronous motor 1 on the basis of the speed response teacher data and the speed response test data.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a machine learning device that determines the responsiveness of a motor control device that performs vector control of a permanent magnet synchronous motor, corrects constants and control gains, and a motor control device that is connected to the machine learning device.

A general speed control system for a permanent magnet synchronous motor is realized by a cascade configuration of two feedback loops, a current minor loop that feedback-controls the torque current and a speed major loop that feedback-controls the motor rotation speed. This configuration can be treated as an independent control system by making the response of the current loop sufficiently high with respect to the response of the velocity loop in order to avoid interference between the responsiveness of the velocity and the torque current.

A PI control system (Proportional Integral Differential Control) is widely used as a control for feedback control of speed and current. At this time, the current loop response is expressed by a second-order lag system in which the resistance, the q-axis inductance, the current proportional gain, and the current integral gain are constants. The velocity loop response is expressed by a quadratic lag system with the motor axis inertia, pole pair number, induced voltage constant, velocity proportional gain, and velocity integration gain as constants when the friction of the mechanical system connected to the permanent magnet synchronous motor is small. Will be done.

In the second-order lag system, both the overshoot characteristic and the response frequency need to be adjusted using two parameters, the damping coefficient and the natural frequency. Therefore, in the second-order lag system, unlike the first-order lag system, the time constant cannot be directly specified to determine the responsiveness. Therefore, in the second-order lag system, it is generally expressed by approximating the first-order lag system in which the response time of the closed loop is set as a time constant. At this time, the proportional gain of the current control system is represented by the product of the time constant of the closed loop and the q-axis inductance, and the integrated gain is represented by the product of the time constant of the closed loop and the resistance.

However, as the error of the q-axis inductance set in the current loop increases with respect to the true value, an error also occurs with respect to the set time constant during the transient response. In particular, when the true value drops with respect to the q-axis inductance set value due to demagnetization or the like, an overcurrent due to overshoot occurs.

Further, as the error becomes larger with respect to the true value of the resistance, an error occurs in the time from the time constant to the steady state. In particular, when the resistance value rises with respect to the DC resistance value due to the skin effect or temperature rise, the time required to reach the steady value after the transient fluctuation becomes long, causing a decrease in the output torque immediately after the transient response. ..

Further, as the error in the set value of the motor shaft inertia set in the speed loop becomes larger with respect to the true value of the inertia due to the inertia fluctuation of the mechanical system, an error occurs with respect to the set time constant. In particular, when the inertia suddenly increases due to an increase in the weight of the mechanical system or the like, the generation of the q-axis current command value that determines the output torque is delayed, and in some cases, step-out occurs.

Therefore, there is a resistance identification method that identifies the true value of the resistance and reflects it in the resistance set value (see, for example, Patent Document 1). This method uses the voltage command, command current, and detected current before resistance value correction when the current command value is switched when positioning the rotor of the permanent magnet synchronous motor, and the current command value and the resistance value of the detected current. The resistance value is identified so that the deviation of the error of is zero.

There is also a method for identifying the q-axis inductance (see, for example, Patent Document 2). In this method, when the permanent magnet synchronous motor is stopped, an alternating voltage is applied to the dq axis to obtain an instantaneous value of each axis current, and the interlinkage magnetic flux obtained by time-integrating the voltage is the vertical axis and the measured axis current. A hysteresis curve is drawn with the horizontal axis as the horizontal axis, and the q-axis inductance is identified from the inclination at the intersection of the center line of the hysteresis curve with the vertical axis.

Japanese Unexamined Patent Publication No. 2007-228767 Japanese Unexamined Patent Publication No. 2003-111499

The above-mentioned resistance identification method shows a method of identifying a resistance value against a change in the resistance value due to aged deterioration or a temperature rise, and obtaining a response according to a predetermined current control response angular frequency. However, if the true resistance value is an abnormal value and the control gain is corrected by this identification operation, there is a problem that the response of the current loop also becomes abnormal.

Further, in the above-mentioned method for identifying the q-axis inductance, a pulse voltage alternating with the d-axis or the q-axis is applied, the time integrated value of the voltage and the current value at that time are acquired and stored by the measuring instrument, and the vertical axis is displayed. This is a method of measuring the d-axis or q-axis inductance from the hysteresis curve when the time-integrated value of voltage and the current value are plotted on the horizontal axis. However, when this method is executed in an embedded microcomputer instead of a measuring instrument, the memory capacity that can store all the time-series data of the voltage and current in the pulse voltage application, or all the acquired time-series data, It is necessary to secure a buffer memory capacity for communicating with an external storage device, and there is a difficult problem to realize with a limited memory capacity.

The machine learning device in the present disclosure is a machine learning device that performs machine learning of a motor control device by connecting a motor that is a permanent magnet synchronous motor to a motor control device that is controlled by vector control.

The machine learning device includes a speed response time series data storage unit that stores response time series data of the rotation speed of the motor, and a speed time constant detection unit that detects the time constant of speed control from the response time series data of the rotation speed. ..

In addition, the machine learning device has an inertia constant correction unit that corrects the inertia constant of the motor shaft of the motor based on the time constant output by the speed time constant detection unit, and an inertia constant and an inertia constant correction unit that is output by the preset inertia constant correction unit. Speed response data generator that generates speed response teacher data and speed response test data based on the correction value, and speed response classification that classifies the speed response of the motor based on the speed response teacher data and speed response test data. It has a part.

Further, the machine learning device in the present disclosure is a machine learning device that performs machine learning of a motor control device by connecting to a motor control device that controls a motor that is a permanent magnet synchronous motor by vector control.

The machine learning device has a current response time-series data storage unit that stores the response time-series data of the dq coordinate current of the motor, and a current time-constant detector that detects the time constant of current control from the time-series data of the dq coordinate current. Be prepared.

In addition, the machine learning device has a motor constant correction unit that corrects the resistance value and q-axis inductance of the motor based on the time constant output by the current time constant detection unit, and a preset motor constant setting value and motor constant correction. The current response data generation unit that generates current response teacher data and current response test data based on the motor constant correction value output by the unit, and the current response class of the motor based on the current response teacher data and current response test data. It is provided with a current response classification unit for classification.

Further, the motor control device in the present disclosure is a motor control device that controls a motor, which is a permanent magnet synchronous motor, by vector control in response to the output from the above-mentioned machine learning device.

The motor control device synchronizes the three-phase current with the current detection unit that detects the three-phase current of the U-phase, V-phase, and W-phase of the stationary coordinate system of the motor, and the rotation position detection unit that detects the actual rotation position of the motor. At least the d-axis current or q-axis current based on the three-phase / dq coordinate conversion unit that converts the d-axis current and q-axis current of the rotation coordinate system, and the rotation speed calculated from the motor speed command value and the actual rotation position. It is provided with a speed control unit that calculates one command current.

In addition, the motor control device has a current control unit that calculates at least one d-axis or q-axis voltage command based on the command current and the d-axis current or q-axis current, and a three-phase voltage command in a synchronous rotation coordinate system. It includes a dq coordinate / three-phase conversion unit that converts to a voltage command, and a PWM control unit that controls the pulse width of the switching element provided in the inverter based on the three-phase voltage command and drives the motor.

In addition, the motor control device includes an inertia constant changing unit that changes the motor shaft inertia constant of the motor, a speed control parameter changing unit that changes the parameters of the speed control unit, and a resistance constant, d-axis inductance constant, and q-axis of the motor. Whether or not at least one of the motor constant change unit for changing the inductance constant, the current control parameter change unit for changing the parameters of the current control unit, the speed control parameter change unit, the motor constant change unit, and the current control parameter change unit can be executed. It is provided with a changeability determination unit that determines.

Further, the motor control device in the present disclosure is a motor control device that controls a motor, which is a permanent magnet synchronous motor, by sensorless vector control in response to the output from the above-mentioned machine learning device.

The motor control device converts the three-phase current into a γ-axis current and a δ-axis current in the estimated rotational coordinate system, and a current detector that detects the three-phase currents of the U-phase, V-phase, and W-phase of the quiescent coordinate system of the motor. Three-phase / γδ coordinate conversion unit, an induced voltage observer unit that calculates at least one velocity-induced voltage of γ-axis velocity-induced voltage or δ-axis velocity-induced voltage based on the γ-axis current and δ-axis current, and a velocity-induced voltage. It is provided with a rotation position estimation unit that calculates the estimated rotation position of the motor based on the above, and a rotation speed estimation unit that calculates the estimated rotation speed of the motor based on the speed-induced voltage.

In addition, the motor control device includes an estimated speed control unit that calculates at least one command current of γ-axis current or δ-axis current based on the speed command value of the motor and the estimated rotation speed calculated by the rotation speed estimation unit, and a command. A γδ-axis current control unit that calculates at least one estimated voltage command value on the γ-axis or δ-axis based on the current and the γδ-axis current, and γδ that converts the estimated voltage command value into a three-phase voltage command value in the estimated rotational coordinate system. It includes a coordinate / three-phase conversion unit and a PWM control unit that controls the pulse width of the switching element provided in the inverter based on the three-phase voltage command value and drives the motor.

In addition, the motor control device includes an estimated speed control parameter changing unit that changes the parameters of the estimated speed control unit, a γδ-axis current control parameter changing unit that changes the parameters of the γδ-axis current control unit, and a motor shaft inertia constant of the motor. The inertia constant change part that changes the motor resistance constant, the d-axis inductance constant, and the q-axis inductance constant, the estimated speed control parameter change part, the motor constant change part, and the γδ-axis current control. It is provided with a changeability determination unit that determines at least one executionability of the parameter change unit.

The machine learning device in the present disclosure and the motor control device connected to the machine learning device perform set value correction and control gain correction in the vector control system of the permanent magnet synchronous motor, and perform highly accurate speed control and current control. To realize.

Configuration diagram of the motor control device and the machine learning device in the first embodiment Closed-loop block diagram of a vector control system composed of a speed control unit and a current control unit according to the first embodiment. Closed-loop block diagram of the current control unit according to the first embodiment The figure which showed an example of the step response of the current control system when there is an error in the set value of resistance with respect to the true value (Ra) of resistance in Embodiment 1. FIG. The figure which showed an example of the step response of the current control system when there is an error in the set value of the q-axis inductance with respect to the true value (Lq) of the q-axis inductance in Embodiment 1. The figure which shows the classification of the current response time series data in Embodiment 1. Closed-loop block diagram of the speed control unit according to the first embodiment The figure which showed an example of the step response of the speed control system when there is an error in the set value of the motor shaft inertia with respect to the true value (J) of the motor shaft inertia in Embodiment 1. The figure which shows the classification of the speed response time series data in Embodiment 1. Configuration diagram of the motor control device and the machine learning device in the second embodiment The figure which shows an example of the phase disturbance response of the estimated phase system in Embodiment 2.

Hereinafter, embodiments will be described in detail with reference to the drawings. However, more detailed explanation than necessary may be omitted. For example, detailed explanations of already well-known matters or duplicate explanations for substantially the same configuration may be omitted.

It should be noted that the accompanying drawings and the following description are provided for those skilled in the art to fully understand the present disclosure, and are not intended to limit the subject matter described in the claims.

(Embodiment 1)
Hereinafter, the first embodiment will be described with reference to FIGS. 1 to 9. FIG. 1 shows the configuration of the motor control device 100 and the machine learning device 101 connected to the motor control device 100.

The motor control device 100 controls the permanent magnet synchronous motor 1 (hereinafter referred to as the motor 1) by vector control according to the output from the machine learning device 101.

The motor control device 100 having a configuration with a position sensor includes a current detection unit 3 that detects three-phase currents of U-phase, V-phase, and W-phase in the stationary coordinate system of the motor 1, and rotation that detects the actual rotation position of the motor 1. A position detection unit 2 is provided.

Further, the motor control device 100 calculates from the three-phase / dq coordinate conversion unit 4 that converts the three-phase current into the d-axis current and the q-axis current of the synchronous rotation coordinate system, the speed command value of the motor 1, and the actual rotation position. A speed control unit 8 that calculates at least one command current of d-axis current or q-axis current based on the rotation speed, and at least one of d-axis or q-axis based on the command current and d-axis current or q-axis current. It includes a current control unit 9 that calculates one voltage command, and a dq coordinate / three-phase conversion unit 10 that converts the voltage command into a three-phase voltage command of a synchronous rotation coordinate system.

Further, although not shown, the motor control device 100 changes the PWM control unit 11 that controls the pulse width of the switching element provided in the inverter to drive the motor 1 and the motor shaft inertia constant of the motor 1 based on the three-phase voltage command. The inertia constant changing unit 15 for changing the inertia constant, the speed control parameter changing unit 14 for changing the parameters of the speed control unit 8, and the motor constant changing unit 13 for changing the resistance constant, the d-axis inductance constant, and the q-axis inductance constant of the motor 1. , A changeability determination that determines at least one of the current control parameter change unit 12 that changes the parameters of the current control unit 9, the speed control parameter change unit 14, the motor constant change unit 13, and the current control parameter change unit 12. A unit 16 is provided.

Here, the rotation speed of the motor 1 is calculated by using the differential processing unit 5 to calculate the amount of change in the rotation position detected by the rotation position detection unit 2. The speed control unit 8 takes a deviation between the detected rotation speed ω and the speed command value ω * of the motor 1, and outputs the q-axis current command value so that the speed deviation becomes zero. The output q-axis current command value and the current detection values of the phase currents Iu, Iv, and Iw of the stationary coordinate system flowing through the motor 1 are determined by the three-phase / dq coordinate conversion unit 4 to determine the dq-axis current id of the synchronous rotating coordinate system. Converted to iq.

The current control unit 9 takes a deviation between the converted dq-axis current id and iq and the current command values id * and iq *, and outputs vd ** and vq ** so that the current deviation becomes zero. In vd ** and vq **, the velocity electromotive force terms interfere with each other between the d-axis and the q-axis, so the sum of the output of the non-interference control unit 6 is used as the voltage command value of the dq-axis. Thereby, the impedance of each axis can be treated as an RL circuit composed of the resistance and the inductance of each axis.

The dq coordinate / three-phase conversion unit 10 converts the voltage command value of the dq axis into the phase voltages Vu, Vv, and Vw of the rest coordinate system. The phase voltages Vu, Vv, and Vw generate a gate signal having a predetermined duty ratio by comparing with a carrier wave of a triangular wave or a sawtooth wave in the PWM control unit 11 (Pulse Width Modulation). The power element of the inverter circuit is driven by this gate signal, and a voltage is applied to each phase of the motor 1.

Here, when the speed command value ω * is set to zero, id * is set to a predetermined value, and iq * is set to zero, a specific rotation position is given and the magnetic pole position of the motor 1 is pulled into a specific rotation position for positioning. The mode for performing the above is set as the positioning mode. After this positioning mode, when the current command id * or iq * is changed in steps or pulses between zero and a predetermined value, the time series data at the time of transient response of the detected current id or iq is processed by the microcomputer communication. It is transmitted from the unit 27 to the machine learning device 101.

That is, the machine learning device 101 is connected to the motor control device 100 that controls the motor 1 to perform machine learning of the motor control device 100.

The machine learning device 101 includes a speed response time series data storage unit 19 that stores the response time series data of the rotation speed of the motor 1, and a speed time constant detection unit that detects the time constant of speed control from the response time series data of the rotation speed. 18 and an inertia constant correction unit 17 that corrects the inertia constant of the motor shaft of the motor 1 based on the time constant output by the speed time constant detection unit 18.

Further, the machine learning device 101 includes a speed response data generation unit 20 that generates speed response teacher data and speed response test data based on a preset inertia constant and an inertia correction value output by the inertia constant correction unit 17. A speed response classification unit 21 for classifying the speed response of the motor 1 based on the speed response teacher data and the speed response test data is provided. Here, the classification refers to classifying the feature vector generated from the speed response test data based on the class label of the speed response teacher data.

Further, the machine learning device 101 detects the time constant of the current control from the current response time-series data storage unit 24 that stores the response time-series data of the dq coordinate current and the voltage of the motor 1 and the time-series data of the dq coordinate current. It includes a current time constant detection unit 23 and a motor constant correction unit 22 that corrects the resistance value of the motor 1 and the q-axis inductance based on the time constant output by the current time constant detection unit 23.

Further, the machine learning device 101 is a current response data generation unit that generates current response teacher data and current response test data based on preset motor constant setting values and motor constant correction values output by the motor constant correction unit 22. 25, and a current response classification unit 26 that classifies the current response of the motor 1 based on the current response teacher data and the current response test data. At this time, the classification refers to classifying the feature vector generated from the current response test data based on the class label of the current response teacher data.

Here, the current response time-series data storage unit 24 and the speed response time-series data storage unit 19 store the data received by the external terminal communication processing unit 28. The current time constant detection unit 23 detects the closed loop time constant of the current control system from the rising waveform of the time series data of the detected current id or iq at the time of transient response. Further, the speed time constant detection unit 18 detects the closed loop time constant of the speed control system from the rising waveform of the time series data of the rotation speed detected at the time of the transient response.

The motor constant correction unit 22 identifies the resistance, d-axis inductance, and q-axis inductance of the motor 1 from the detected time constant, and transmits the identification result from the external terminal communication processing unit 28 to the motor control device 100. Further, the inertia constant correction unit 17 identifies the inertia in the motor shaft of the motor 1 from the detected time constant, and transmits the identification result from the external terminal communication processing unit 28 to the motor control device 100.

At this time, the identification result data is not limited to communication applications, and the identification result can be confirmed by the user by outputting it as display data to a display device (not shown) such as an LCD (Liquid Crystal Display).

On the other hand, the current response data generation unit 25 receives a current response from the feature quantity vector of the closed loop step response of the motor model of the first-order lag system and the current response class label CR _labels at a predetermined motor constant set value. The teacher data D _{cu_train} is generated, and the current response test data D _{cu_test} is generated from the current response time series data. The current response teacher data D _{cu_train} and the current response test data D _{cu_test} are composed of a feature vector and a class label. The feature amount vector includes at least one of three, an overshoot amount, a steady-state deviation value, and a time constant. When the feature vector is expressed in tabular form, there is one class vector for each row of the feature vector.

Further, the speed response data generation unit 20 includes a feature quantity vector of the closed loop step response of the mechanical system model of the integral system or the first-order lag system at a predetermined motor constant set value, and the speed response class label CR _labels . From, the speed response teacher data D _{sp_train} is generated, and the speed response test data D _{sp_test} is generated from the speed response time series data. The velocity response teacher data D _{sp_train} and the velocity response test data D _{sp_test} are composed of a feature vector and a class label. The feature amount vector includes at least one of three, an overshoot amount, a steady-state deviation value, and a time constant. When the feature vector is expressed in tabular form, there is one class vector for each row of the feature vector.

A method of detecting each feature vector from the current response time series data will be described. First, the current response data generation unit 25 slices only the response data from the time series data. The start trigger for slicing by the current response data generation unit 25 is performed at the timing when the current command value changes. If the time width for slicing by the current response data generation unit 25 is about 100 times the current response frequency, the steady state can be sufficiently confirmed.

In this way, the overshoot amount can be obtained by detecting the maximum value in the sliced data. Further, the value of the steady deviation can be obtained by taking the difference between the average value of an arbitrary number of samples and the current command value from the tail of the sliced data. Further, the time constant can be obtained by calculating the time from the time when the current command value changes to the time when the detected current reaches 0.632 times the current command value after the transition by multiplying the number of samples and the sample period. However, the sample period needs to be sufficiently short with respect to the time constant of the current response.

As described above, the feature amount vector of each response can be acquired from the time series data including the response data of a plurality of times, and the current response test data D _{cu_test} can be generated.

On the other hand, a method of detecting each feature vector from the velocity response time series data will be described. First, the speed response data generation unit 20 slices only the response data from the time series data. The start trigger for slicing by the speed response data generation unit 20 is performed at the timing when the speed command value changes. If the time width for slicing by the speed response data generation unit 20 is about 100 times the speed response frequency, the steady state can be sufficiently confirmed.

In this way, the overshoot amount can be obtained by detecting the maximum value in the sliced data. Further, the value of the steady deviation can be obtained by taking the difference between the average value of an arbitrary number of samples and the velocity command value from the tail of the sliced data. The time constant is calculated by multiplying the number of samples by the sample period from the time when the speed command value changes to the time when the detected rotation speed reaches 0.632 times the speed command value after the transition. can. However, the sample period needs to be sufficiently short with respect to the time constant of the velocity response.

As described above, the feature amount vector of each response can be acquired from the time series data including the response data of a plurality of times, and the velocity response test data D _{sp_test} can be generated.

The current response classification unit 26 classifies the current response test data D _{cu_test} based on the class label of the current response teacher data D _{cu_train} , and transmits the classification result from the external terminal communication processing unit 28 to the motor control device 100. Similarly, the speed response classification unit 21 classifies the speed response test data D _{sp_test} based on the class label of the speed response teacher data D _{sp_train} , and transmits the classification result from the external terminal communication processing unit 28 to the motor control device 100. The machine learning method of the classification algorithm in the current response classification unit 26 and the speed response classification unit 21 is not particularly limited to a multi-layer perceptron, a random forest, a naive bays, and the like. Here, the class label is a label indicating the classification result determined in advance based on the statistical probability of the past judgment result data group, the judgment result of an expert, etc., and the number of classes is the type of class label. It is a number.

The microcomputer communication processing unit 27 of the motor control device 100 receives the identification result of the motor constant and the class label of the current response or the class label of the speed response, and in the changeability determination unit 16, the speed control parameter change unit 14 and the motor constant change. It is determined whether or not at least one of the unit 13 and the current control parameter changing unit 12 can be executed. That is, the changeability determination unit 16 determines whether or not the speed control unit 8 is changed by the speed control parameter change unit 14. Further, the changeability determination unit 16 determines whether or not the non-interference control unit 6 is changed in the motor constant change unit 13. Further, the current control parameter changing unit 12 determines whether or not to change the current control unit 9.

When executing the change, the speed control parameter changing unit 14 calculates the control gain based on the identification result of the inertia constant obtained from the inertia constant changing unit 15, and overwrites the control gain of the speed control unit 8. .. Further, the motor constant changing unit 13 overwrites the motor constant used by the non-interference control unit 6 based on the identification result of the motor constant. Further, the current control parameter changing unit 12 calculates the control gain based on the identification result of the motor constant obtained from the motor constant changing unit 13, and overwrites the control gain of the current control unit 9.

Here, the number of classes is 2 classes, with the normal region being the one in which the current response time series data fits within the range of the maximum to minimum values assumed for resistance, d-axis inductance, and q-axis inductance, and the other being the abnormal region. Divide into. By executing the above change in the normal area and not executing the above change in the abnormal area, it is possible to determine whether or not the change is possible. However, the number of classes is not limited to two classes, and may be set to two or more classes. For example, by further adding another current threshold value or another time threshold value, dividing the area, and assigning a new class label to the added area, the number of classes can be easily increased to 3 or more.

Here, "the principle of identifying the motor constant (resistance, d-axis inductance, q-axis inductance) and changing the control gain of the current control unit) and" the principle of identifying the inertia constant and changing the control gain of the speed control unit ", respectively. explain.

First, "the principle of identifying motor constants (resistance, d-axis inductance, q-axis inductance) and changing the control gain of the current control unit" will be described below.

FIG. 2 shows a closed-loop block diagram of a vector control system composed of a speed control unit and a current control unit. The vector control system is composed of a current minor loop composed of a current PI control unit 43 and a motor model 44, and a cascade connection of a speed outer loop composed of a speed PI control unit 42 and a mechanical system model 45. When the response frequency of the current control system is sufficiently higher than the response frequency of the speed control system, it can be analyzed as an independent control system. (Equation 1) shows the current control system closed-loop transfer function of the motor when PI control is used for the controller.

Ra in (Equation 1) is the resistance value, Lq is the q-axis inductance, Kpq is the proportional gain of the current control system, and Kiq is the integrated gain of the current control system. The block diagram of (Equation 1) is composed of a current PI control unit 43 and a motor model 44, and is represented as a closed loop block diagram of the current control unit shown in FIG. The closed-loop block diagram of the current control unit includes a step command unit 41 that changes the current command in steps, a current PI control unit 43 that controls the current deviation to be zero, and a motor model 44.

(Equation 1) is a second-order lag system in which the zero point is −Kiq / Kpq, and as it is, the response frequency cannot be specified by one variable like the first-order lag system. Therefore, by arranging the poles as -Kiq / Kpq = -Ra / Lq so as to cancel the zeros and poles of the transfer function, the first-order transfer function of (Equation 2) is obtained, and at this time, the proportional gain is (Equation 3), the integrated gain is determined as in (Equation 4).

Ωc in (Equation 2) to (Equation 4) is the response frequency of the current control system closed loop, and is equal to Kpq / Lq. Further, when (Equation 2) is established, overshoot and vibration as seen in the secondary lag system do not occur.

At this time, after the mode in which the speed command value ω * is set to zero, the magnetic pole position is pulled to a specific rotation position, and the motor rotor is positioned, the rise when the q-axis current command value is changed in a step shape or a pulse shape. The response of time is as shown in (Equation 5) in the expression by the transfer function, and the time-series data follows the current command value by the function as in (Equation 6).

Iqo in (Equation 5) and (Equation 6) is a target q-axis current command value. By determining that Iqo has a torque smaller than the starting torque of the load, the response can be confirmed while the rotor is stopped. As for the response of the d-axis current at the rising edge, the same characteristics as those of the q-axis can be obtained.

However, if the set value of the motor constant has an error from the true value, the canceling condition between the zero point and the pole in (Equation 1) is not strictly satisfied, and (Equation 2) is not satisfied.

For example, FIG. 4 shows a step response waveform of the current control system when there is an error between the set value of the resistance value and the true value. FIG. 4 is an example of the step response of the current control system when there is an error in the set value of the resistance with respect to the true value (Ra) of the resistance in the first embodiment. In FIG. 4, the horizontal axis represents time, and the vertical axis represents the q-axis current of the synchronous rotating coordinate system at the time of step response. The set value of the resistance is in the range of 0.1Ra to 10Ra.

In the step response waveform of FIG. 4, the responsiveness in the first half of the rise is almost the same waveform regardless of the set value of the resistance value (true value and error), and the error accuracy of the set value of the q-axis inductance is dominant. You can see that. On the other hand, the waveforms after the time constant have a remarkable difference. If the resistance value is larger than the resistance value of the nominal model, the time required to reach the current command value is prolonged, which causes a decrease in output torque.

Further, FIG. 5 shows a step response waveform of the current control system when there is an error between the set value and the true value of the q-axis inductance. FIG. 5 is an example of the step response of the current control system when there is an error in the set value of the q-axis inductance with respect to the true value (Lq) of the q-axis inductance in the first embodiment. It is a closed loop response waveform of the current control system when the q-axis inductance is swept between 0.1 and 10 times the resistance value of the nominal model, especially when the q-axis inductance is larger than the resistance value of the nominal model. , Overshoot occurs.

At the above, first, the principle of identifying the true value of the q-axis inductance is shown below. The actual response frequency to the set value of the q-axis inductance is (Equation 7).

Ωsens in (Equation 7) is the actual response frequency observed from the time when the current command value is switched until it reaches 63.2% of the current command value. Kpqset is the proportional gain of the current control system defined by (Equation 3) according to the set values of the response frequency ωc and the q-axis inductance, and Lq is the true value of the q-axis inductance. At this time, since the actual response frequency ωsens was obtained with respect to the target response frequency ωc, the ratio of the response frequencies becomes the ratio of the q-axis inductance from (Equation 3) at this time. The relationship is shown in (Equation 8).

In (Equation 8), (Lq) ^ is the identification value of the q-axis inductance, and Lqset is the set value of the q-axis inductance.

Therefore, the true value of the q-axis inductance can be identified based on (Equation 8), and the proportional gain Kpq of the current control system can be changed from (Equation 3) based on the identification result. The relationship is shown in (Equation 9).

The d-axis inductance and the proportional gain of the current control system can be corrected by the same method as described above.

Next, the principle for identifying the true value of the resistance value is shown below. When there is an error between the true value of the resistance value and the set value, and T1 is the time when the current command value becomes 63.2% of the final value from 0, there is a large difference in the responsiveness after the time T1. At this time, assuming that the time n times the time constant (n is an integer of 1 or more) is nT, when the zero pole cancellation of (Equation 2) is established, (Equation 10) becomes the true value current.

However, when the resistance value is larger than the true value, the time to reach the q-axis current value iq (nT) of (Equation 10) is extended, and when the resistance value is smaller than the true value, the q-axis current value iq (Equation 10) is reached. The time to reach nT) is shortened. From the above, the deviation between the time nT for reaching the q-axis current value of (Equation 10) and the time until the observed value actually reaches the q-axis current value of (Equation 10) becomes zero during the transient response. In addition, by searching for the resistance set value, the identification value (Ra) ^ of the true resistance value can be obtained.

Further, the integrated gain Kic of the current control system can be changed from (Equation 4). The relationship is shown in (Equation 11).

As described above, the identification of the motor constants described above is executed by the machine learning device 101.

FIG. 6 shows the classification of the current response time series data in the first embodiment. In FIG. 6, the horizontal axis represents time and the vertical axis represents the q-axis current in the synchronous rotating coordinate system. The white area in the background indicates a normal area, and the colored area indicates an abnormal area.

The current response classification is based on at least one of the rise time time constant, overshoot amount, and steady deviation values of the response time series data in the dq coordinate system current sent from the motor controller 100. It is executed by the machine learning device 101.

After that, the identification result is sent back to the motor control device 100, and when the response waveform after the identification result falls within the class of the normal region shown in FIG. 6, the motor constants of the current control unit 9 and the non-interference control unit 6 are changed. , The value of the dq axis voltage can be corrected. The relationship regarding the q-axis voltage command is shown in (Equation 12). Ti in (Equation 12) is the integration time.

Next, the "principle of identifying the inertia constant and changing the control gain of the speed control unit" will be described below.

(Equation 13) shows the speed control system closed loop transfer function of the motor 1 when PI control is used for the controller.

In (Equation 13), J is the motor shaft inertia, D is the viscosity coefficient for friction during rotation, Pn is the number of pole pairs of the motor 1, Ke is the induced voltage constant of the motor 1, Kps is the proportional gain of the speed control unit 8, Kis. Is the integrated gain of the speed control unit 8. As a precondition for the establishment of (Equation 13), the response frequency ωc of the current control unit 9 is assumed to be sufficiently larger than the response frequency ωcs of the speed control unit 8 (ωc >> ωcs).

The block diagram of (Equation 13) is composed of a speed PI control unit 42 and a mechanical system model 45, and is represented as a closed loop block diagram of the speed control unit shown in FIG. 7. The closed loop block diagram of the speed control unit includes a step command unit 41 that changes the speed command in steps, a speed PI control unit 42 that controls the speed deviation to be zero, and a mechanical model 45.

(Equation 13) is a second-order lag system in which the zero point is −Kis / Kps, and as it is, the response frequency cannot be specified by one variable like the first-order lag system. Therefore, by arranging the poles as -Kis / Kps = -D / J, Kis = Kpωf so as to cancel the zero point of the transfer function and the pole, the first-order transfer function of (Equation 14) is obtained. At this time, the proportional gain of the speed control unit 8 is determined as (Equation 15), and the integrated gain is determined as (Equation 16).

Ωcs in (Equation 14) to (Equation 16) is the response frequency of the speed control system closed loop, and is equal to Kps / J. ωf is an arbitrary response frequency that is sufficiently small with respect to ωcs. At this time, the speed command value ω * is set to a constant value, and after the motor 1 reaches the synchronous rotation speed, the response at the time of rising when the speed command value is changed in a step shape or a pulse shape is expressed by the transmission function. , (Equation 17), and the time-series data follows the speed command value with a function like (Equation 18).

Δωstep in (Equation 17) and (Equation 18) is the difference between the speed command value before the transient response and the speed command value after the transient response. However, if the set value of the inertia constant J has an error from the true value, the canceling condition between the zero point and the pole in (Equation 13) is not strictly satisfied, and (Equation 14) is not satisfied.

FIG. 8 shows an example of the step response of the speed control system when there is an error in the set value of the motor shaft inertia with respect to the true value (J) of the motor shaft inertia in the first embodiment. In FIG. 8, the horizontal axis represents time and the vertical axis represents the rotation speed of the motor 1. The set value of the motor shaft inertia is in the range of 0.1J to 10J.

FIG. 8 shows a closed-loop response waveform of the speed control system when the inertia value is swept between 0.1 and 10 times the inertia value of the nominal model. In particular, the inertia value is higher than the inertia value of the nominal model. If it is large, overshoot occurs and it takes time to reach a desired speed, so that the generation of the q-axis current command is delayed and the torque response is lowered.

In the above cases, the principle for identifying the true value of the motor shaft inertia is shown below. The actual response frequency to the set value of the q-axis inductance is (Equation 19).

Ωsp_sens in (Equation 19) is the observed actual response frequency from the time when the speed command value is switched until it reaches 63.2% of the speed command value. Kps_set is the proportional gain of the speed control system determined by (Equation 15) by the set values of the response frequency ωcs and the motor shaft inertia J, and J is the true value of the motor shaft inertia J. At this time, since the actual response frequency ωsp_sens was obtained with respect to the target response frequency ωcs, the ratio of the response frequencies is the ratio of the motor shaft inertia from (Equation 15). The relationship is shown in (Equation 20).

(J) ^ in (Equation 20) is the identification value of the motor shaft inertia, and Jset is the set value of the motor shaft inertia.

Therefore, the true value of the motor shaft inertia can be identified based on (Equation 20), and based on the identification results, the proportional gains Kps and Kis of the speed control system are obtained from (Equation 15) and (Equation 16). Can be changed. The relationship is shown in (Equation 21) and (Equation 22).

The identification of the motor shaft inertia executed above is executed by the machine learning device 101 based on the response time series data sent from the motor control device 100. The machine learning device 101 sends the identification result back to the motor control device 100.

FIG. 9 shows the classification of the speed response time series data in the first embodiment. In FIG. 9, the horizontal axis represents time and the vertical axis represents the rotation speed of the motor 1. The white area in the background indicates a normal area, and the colored area indicates an abnormal area.

When the response waveform after the identification result falls within the class of the normal region shown in FIG. 9, the control gain of the speed control unit 8 can be changed and the q-axis current command value can be corrected. The relationship between the q-axis voltage command and the control gain of the speed control unit 8 is shown in (Equation 23).

As described above, in the present embodiment, the machine learning device 101 connects the motor 1 which is a permanent magnet synchronous motor to the motor control device 100 which is controlled by vector control, and performs machine learning of the motor control device 100. It is a device.

The machine learning device 101 includes a speed response time series data storage unit 19 that stores response time series data of the rotation speed of the motor 1, and a speed time constant detection unit that detects a speed control time constant from the response time series data of the rotation speed. Equipped with 18.

In addition, the machine learning device 101 includes an inertia constant correction unit 17 that corrects the inertia constant of the motor shaft of the motor 1 based on the time constant output by the speed time constant detection unit 18, and a preset inertia constant and inertia constant correction unit. The speed response data generation unit 20 that generates the speed response teacher data and the speed response test data based on the inertia correction value output by, and the speed response of the motor based on the speed response teacher data and the speed response test data. It is provided with a speed response classification unit 21 for classifying.

With this configuration, in the vector control system of the permanent magnet synchronous motor 1, set value correction and control gain correction are executed, and highly accurate speed control and current control can be realized.

Further, as in the present embodiment, the speed response classification unit 21 determines the value of the time constant at the time of rising, the amount of overshoot, and the value of the steady deviation when the speed command value is given during the synchronous rotation of the motor 1. The classification classified based on at least one may be output to the motor control device 100.

Further, as in the present embodiment, the inertia constant correction unit 17 identifies the inertia value of the motor shaft of the motor 1 from the time constant of the speed time constant detection unit 18, and outputs the output to the motor control device 100 to drive the motor 1. May be controlled.

Further, as in the present embodiment, the velocity response teacher data is the transient response time series data of the integral system or the first-order lag system generated based on the predetermined velocity response frequency and one or more inertia set values at the rising edge. The classification based on the feature amount vector including at least one of the time constant, the overshoot amount, and the steady deviation value of the above may be output to the motor control device 100.

Further, the motor control device 100 of the present embodiment is a motor control device that controls the motor 1 which is a permanent magnet synchronous motor by vector control in response to the output from the machine learning device 101.

The motor control device 100 includes a current detection unit 3 that detects a three-phase current of U-phase, V-phase, and W-phase in the stationary coordinate system of the motor 1, a rotation position detection unit 2 that detects the actual rotation position of the motor 1. Based on the three-phase / dq coordinate conversion unit 4 that converts the three-phase current into the d-axis current and the q-axis current of the synchronous rotation coordinate system, and the rotation speed calculated from the speed command value of the motor 1 and the actual rotation position, d. A speed control unit 8 for calculating at least one command current of the shaft current or the q-axis current is provided.

In addition, the motor control device 100 has a current control unit 9 that calculates at least one voltage command of the d-axis or the q-axis based on the command current and the d-axis current or the q-axis current, and the voltage command is a synchronous rotation coordinate system. The dq coordinate / three-phase conversion unit 10 that converts to the three-phase voltage command of the above, and the PWM control unit 11 that controls the pulse width of the switching element provided in the inverter and drives the motor based on the three-phase voltage command, although not shown. Be prepared.

In addition, the motor control device 100 includes an inductance constant changing unit 15 that changes the motor shaft inductance constant of the motor 1, a speed control parameter changing unit 14 that changes the parameters of the speed control unit 8, and a resistance constant of the motor 1, d. The motor constant changing unit 13 for changing the shaft inductance constant and the q-axis inductance constant, the current control parameter changing unit 12 for changing the parameters of the current control unit 9, the speed control parameter changing unit 14, the motor constant changing unit 13, and the motor constant changing unit 13. The current control parameter changing unit 12 includes at least one changeability determining unit 16 that determines whether or not execution is possible.

With this configuration, in the vector control system of the permanent magnet synchronous motor 1, set value correction and control gain correction are executed, and highly accurate speed control and current control can be realized.

The machine learning device 101 of FIG. 1 learns the speed control system (blocks of inertia constant correction unit 17 to speed response classification unit 21) and blocks the current control system (motor constant correction unit 22 to current response classification unit 26). ) With learning. In the first embodiment, only the speed control system may be used. Further, if the machine learning device 101 learns both the speed control system and the current control system, more accurate speed control and current control can be realized.

(Embodiment 2)
Hereinafter, the second embodiment will be described with reference to FIGS. 10 and 11. FIG. 10 shows the configuration of the motor control device 200 having a position sensorless configuration and the machine learning device 101 connected to the motor control device 200. The machine learning device 101 is connected to the motor control device 200 that controls the motor 1 to perform machine learning of the motor control device 200. Since the machine learning device 101 is the same as the machine learning device 101 of the first embodiment, detailed description thereof will be omitted. The same configurations and operations as those of the first embodiment are given the same numbers as those of the first embodiment, and detailed description thereof is omitted.

The motor control device 200 controls the motor 1 by sensorless vector control according to the output from the machine learning device 101. The motor control device 200 estimates rotational coordinates of the three-phase currents obtained by the current detection unit 3 and the current detection unit 3 that detects the three-phase currents of the U-phase, V-phase, and W-phase of the stationary coordinate system of the motor 1. It is provided with a three-phase / γδ coordinate conversion unit 31 that converts the γ-axis current and the δ-axis current of the system.

Further, the motor control device 200 has an induced voltage observer unit 32 that calculates at least one velocity-induced voltage of the γ-axis velocity-induced voltage or the δ-axis velocity-induced voltage based on the γ-axis current and the δ-axis current, and the velocity-induced voltage. A rotation position estimation unit 33 that calculates an estimated rotation position of the motor 1 based on the above, and a rotation speed estimation unit 34 that calculates an estimated rotation speed of the motor 1 based on the speed-induced voltage are provided.

Further, the motor control device 200 is an estimated speed control unit that calculates at least one command current of the γ-axis current or the δ-axis current based on the speed command value of the motor 1 and the estimated rotation speed calculated by the rotation speed estimation unit 34. 35, the γδ-axis current control unit 36 that calculates at least one estimated voltage command value of the γ-axis or the δ-axis based on the command current and the γδ-axis current, and the three-phase voltage of the estimated rotation coordinate system that calculates the estimated voltage command value. It includes a γδ coordinate / three-phase conversion unit 37 that converts to a command value, and a PWM control unit 11 that controls the pulse width of a switching element provided in the inverter based on the three-phase voltage command value and drives the motor 1.

Further, the motor control device 200 includes an estimated speed control parameter changing unit 38 that changes the parameters of the estimated speed control unit 35, a γδ axis current control parameter changing unit 39 that changes the parameters of the γδ axis current control unit 36, and a motor 1. The inertia constant changing unit 15 for changing the motor axis inertia constant, the motor constant changing unit 13 for changing the resistance constant, d-axis inductance constant, and q-axis inductance constant of the motor 1, the estimated speed control parameter changing unit 38, and the motor. It includes a constant change unit 13 and a changeability determination unit 16 that determines at least one executionability of the γδ-axis current control parameter change unit 39.

In the motor control device 200, the estimated rotation speed of the motor 1 is such that the current detection unit 3 first detects the three-phase current, and the three-phase / γδ coordinate conversion unit 31 uses the three-phase current as the γ-axis current of the estimated rotation coordinate system. Convert to δ-axis current. The induced voltage observer unit 32 calculates the induced voltage from the γ-axis current, the δ-axis current, the motor constant, the γ-axis voltage, and the δ-axis voltage, and the rotation speed estimation unit 34 calculates the estimated rotation speed based on the induced voltage. ..

The estimated speed control unit 35 takes a deviation between the detected estimated rotation speed (ω) ^ and the speed command value ω * of the motor 1, and outputs a δ-axis current command value so that the speed deviation becomes zero. The output δ-axis current command value and the current detection values of the static coordinate system phase currents Iu, Iv, and Iw flowing through the motor 1 are determined by the three-phase / γδ coordinate conversion unit 31 to determine the estimated rotational coordinate system γδ-axis current iγ. Converted to iδ.

The γδ-axis current control unit 36 takes a deviation between the converted γδ-axis currents iγ and iδ and the current command values iγ * and iδ *, and outputs vγ ** and vδ ** so that the current deviation becomes zero. .. In vγ ** and vδ **, the velocity electromotive force terms interfere with each other between the γ axis and the δ axis, so the sum of the output of the non-interference control unit 6 is used as the voltage command value of the γδ axis. Therefore, the impedance of each axis can be treated as an RL circuit composed of a resistance and an inductance of each axis.

The γδ coordinate / three-phase conversion unit 37 converts the voltage command value of the γδ axis into the phase voltages Vu, Vv, and Vw of the rest coordinate system. The phase voltages Vu, Vv, and Vw generate a gate signal having a predetermined duty ratio in the PWM control unit 11 by comparing with a carrier wave of a triangular wave or a sawtooth wave. The power element of the inverter circuit is driven by this gate signal, and a voltage is applied to each phase of the motor 1. However, while the motor 1 is rotating synchronously, the γ axis of the estimated rotating coordinate coincides with the d axis of the synchronous rotating coordinate system, and the δ axis of the estimated rotating coordinate coincides with the q axis of the synchronous rotating coordinate system.

The estimated speed control parameter changing unit 38 changes the control gain of the estimated speed control unit 35 based on the identified motor shaft inertia constant. Further, the γδ-axis current control parameter changing unit 39 changes the control gain of the γδ-axis current control unit 36 based on the identified resistance value, d-axis inductance, and q-axis inductance. The motor constant changing unit 13 changes the motor constant of the non-interference control unit 6 and the motor constant of the induced voltage observer unit 32. The relationship between the motor constant and the induced voltage in the induced voltage observer unit 32 is shown in (Equation 24) and (Equation 25).

In (Equation 24) and (Equation 25), (eγ) ^ is the γ-axis induced voltage obtained by the induced voltage observer unit 32, and (eδ) ^ is the δ-axis induced voltage obtained by the induced voltage observer unit 32. , Vγ is γ-axis voltage, Vδ is δ-axis voltage, (Ra) ^ is resistance identification value, iγ is γ-axis current, iδ is δ-axis current, ω is estimated rotation speed, (Lq) ^ is q-axis inductance identification value , (Ld) ^ are d-axis inductance identification values. At this time, the estimated phase error between the d-axis and the γ-axis is (Equation 26) from the relationship between the induced voltage of the γδ-axis and the estimated phase error Δθ in the estimated rotating coordinate system.

From the estimated position error obtained in (Equation 26), the rotation speed estimation unit 34 can calculate the estimated rotation speed using (Equation 27). Further, the rotation position estimation unit 33 can calculate the estimated rotation position using (Equation 28) from the estimated rotation speed obtained by (Equation 27). Kpes in (Equation 27) is the proportional gain of the rotation speed estimation unit 34, and Kies is the integral gain of the rotation speed estimation unit 34.

From the above, accurate γδ-axis induced voltage can be calculated by changing the motor constant identification value closer to the true value in (Equation 24) and (Equation 25) in the induced voltage observer, and as a result, the estimated rotation speed (Estimated rotation speed) The estimation accuracy of ω) ^ and the estimated rotation position (θ) ^ can be improved.

FIG. 11 shows an example of the phase disturbance response of the estimated phase system in the second embodiment. In FIG. 11, the horizontal axis indicates time, and the vertical axis indicates the phase deviation between the actual phase and the estimated phase.

Quantitative evaluation of the responsiveness of the estimated phase system during motor rotation is difficult with step response because the phase changes continuously. For example, in an environment where a rotation position sensor can be temporarily attached, by applying a disturbance to the estimated phase and adjusting the phase deviation setting time so that the estimated phase follows the sensor phase, as shown in FIG. , Kpes, Kies can be determined.

As described above, in the present embodiment, the machine learning device 101 connects the motor 1 which is a permanent magnet synchronous motor to the motor control device 200 which is controlled by vector control, and performs machine learning of the motor control device 200. It is a device.

The machine learning device 101 has a current response time-series data storage unit 24 that stores the response time-series data of the dq coordinate current of the motor 1, and a current time constant that detects the time constant of current control from the time-series data of the dq coordinate current. A detection unit 23 is provided.

In addition, the machine learning device 101 has a motor constant correction unit 22 that corrects the resistance value of the motor 1 and the q-axis inductance based on the time constant output by the current time constant detection unit 23, and a preset motor constant setting. Based on the value and the motor constant correction value output by the motor constant correction unit 22, the current response data generation unit 25 that generates the current response teacher data and the current response test data, and the current response teacher data and the current response test data. , A current response classification unit 26 for classifying the current responsiveness of the motor 1 is provided.

With this configuration, in the vector control system of the permanent magnet synchronous motor 1, set value correction and control gain correction are executed, and highly accurate speed control and current control can be realized.

Further, as in the present embodiment, the current response classification unit 26 sets the dq coordinate system with respect to the time series data of the dq coordinate system current when the current command value is given in the positioning of the rotor of the motor 1. The classification based on at least one of the time constant of the rise time of the current, the amount of overshoot, and the value of the steady deviation may be output to the motor control device 200.

Further, as in the present embodiment, the motor constant correction unit sets at least one of the true resistance value of the motor 1, the true inductance value of the d-axis of the motor 1, and the true inductance value of the q-axis of the motor 1 as a current. It may be identified from the time constant of the time constant detection unit 23 and output to the motor control device 200.

Further, as in the present embodiment, the current response teacher data is a transient response of an integral system or a first-order lag system generated based on at least one of a resistance set value, a d-axis inductance set value, and a q-axis inductance set value. The motor control device 200 classifies the classification based on the feature quantity vector including at least one of the rising time constant, the overshoot amount, and the steady deviation value in the time series data, and the predetermined current response frequency. It may be output to.

Further, as in the present embodiment, the motor control device 200 may control the motor 1 by sensorless vector control.

As a result, in a motor control device having a position sensorless configuration, there is an effect that step-out is less likely to occur due to a change in the estimated phase portion based on the motor parameter identification result.

Further, the motor control device 200 in the present embodiment is a motor control device that controls the motor 1 which is a permanent magnet synchronous motor by sensorless vector control in response to the output from the machine learning device 101.

The motor control device 200 includes a current detection unit 3 for detecting the three-phase currents of the U-phase, V-phase, and W-phase of the stationary coordinate system of the motor 1, and the three-phase current, and the γ-axis current and the δ-axis of the estimated rotation coordinate system. An induced voltage observer that calculates at least one velocity-induced voltage of γ-axis velocity-induced voltage or δ-axis velocity-induced voltage based on the three-phase / γδ coordinate conversion unit 31 that converts to current and the γ-axis current and δ-axis current. The unit 32 includes a rotation position estimation unit 33 that calculates an estimated rotation position of the motor based on the speed-induced voltage, and a rotation speed estimation unit 34 that calculates the estimated rotation speed of the motor based on the speed-induced voltage.

In addition, the motor control device 200 calculates an estimated speed control that calculates at least one command current of the γ-axis current or the δ-axis current based on the speed command value of the motor 1 and the estimated rotation speed calculated by the rotation speed estimation unit 34. Three phases of the unit 35, the γδ-axis current control unit 36 that calculates at least one estimated voltage command value of the γ-axis or the δ-axis based on the command current and the γδ-axis current, and the estimated voltage command value of the estimated rotational coordinate system. It includes a γδ coordinate / three-phase conversion unit 37 that converts into a voltage command value, and a PWM control unit 11 that controls the pulse width of a switching element provided in the inverter based on the three-phase voltage command value and drives the motor 1.

In addition, the motor control device 200 includes an estimated speed control parameter changing unit 38 that changes the parameters of the estimated speed control unit 35, a γδ axis current control parameter changing unit 39 that changes the parameters of the γδ axis current control unit, and a motor 1. The inertia constant changing unit 15 for changing the motor axis inertia constant, the motor constant changing unit 13 for changing the resistance constant, d-axis inductance constant, and q-axis inductance constant of the motor 1, the estimated speed control parameter changing unit 38, and the motor. It includes a constant change unit 13 and a changeability determination unit 16 that determines at least one executionability of the γδ-axis current control parameter change unit 39.

With this configuration, in the vector control system of the permanent magnet synchronous motor 1, set value correction and control gain correction are executed, and highly accurate speed control and current control can be realized.

The machine learning device 101 of FIG. 10 learns the speed control system (blocks of inertia constant correction unit 17 to speed response classification unit 21) and blocks the current control system (motor constant correction unit 22 to current response classification unit 26). ) With learning. In the second embodiment, only the current control system may be used. Further, if the machine learning device 101 learns both the speed control system and the current control system, more accurate speed control and current control can be realized.

(Other embodiments)
As described above, Embodiments 1 and 2 have been described as examples of the techniques disclosed in the present application. However, since the above-described embodiment is for exemplifying the technique in the present disclosure, the present invention is not limited to this, and changes, replacements, additions, omissions, and the like can be made. Further, it is also possible to combine the components described in the first and second embodiments to form a new embodiment.

In the first and second embodiments, the machine learning device 101 is configured to be easily connected at the time of adjustment such as manufacturing or repair of the motor control devices 100 and 200. Although the machine learning device 101 and the motor control devices 200 and 200 are described as separate bodies, the machine learning device 101 may be mounted on a product (for example, an electric washing machine) together with the motor control devices 100 and 200 and integrated. good. Alternatively, the machine learning device 101 may be configured to be easily connected to the motor control devices 100 and 200 via a communication line such as a network, and may perform machine learning remotely.

Since the above-described embodiment is for exemplifying the technique in the present disclosure, various changes, replacements, additions, omissions, etc. can be made within the scope of claims or the equivalent thereof.

Further, any combination of the above components, a conversion of the expression of the present disclosure between methods, devices, systems, recording media, computer programs and the like is also effective as an aspect of the present disclosure.

In the present disclosure, the control performance is deteriorated by correcting the control gain of the speed control system or the current control system in the vector control algorithm from the response time series data, and correcting the motor constants of the non-interference control unit and the induced voltage observer unit. Can be prevented.

Specifically, it can be applied to satisfy the torque response requirement in a vertical type washing machine or a drum type washing machine in which the motor shaft inertia and the required load torque greatly change in each washing process. The washing machine may include a washer / dryer.

Further, it can be applied to improve the estimation accuracy of the induced voltage observer in the compressor motor of the air conditioner outdoor unit whose winding resistance value fluctuates due to a wide range of temperature changes.

In addition, by constantly sucking response time-series data as IoT (Internet of Things) data to a cloud or on-premises data center, it becomes data to judge the abnormal state and aging deterioration of the device, so it is a remote device. It can be used for services such as condition diagnosis and smooth repair response.

1 Permanent magnet synchronous motor (motor)
2 Rotation position detection unit 3 Current detection unit 4 Three-phase / dq coordinate conversion unit 5 Differentiation processing unit 6 Non-interference control unit 8 Speed control unit 9 Current control unit 10 dq coordinate / three-phase conversion unit 11 PWM control unit 12 Current control parameter Change unit 13 Motor constant change unit 14 Speed control parameter change unit 15 Inertia constant change unit 16 Changeability determination unit 17 Initiality constant correction unit 18 Speed time constant detection unit 19 Speed response time series data storage unit 20 Speed response data generation unit 21 Speed Response classification unit 22 Motor constant correction unit 23 Current time constant detection unit 24 Current response time series data storage unit 25 Current response data generation unit 26 Current response classification unit 27 Microcomputer communication processing unit 28 External terminal communication processing unit 31 Three-phase / γδ coordinates Conversion unit 32 Induced voltage observer unit 33 Rotation position estimation unit 34 Rotation speed estimation unit 35 Estimated speed control unit 36 γδ axis current control unit 37 γδ coordinate / three-phase conversion unit 38 Estimated speed control parameter change unit 39 γδ axis current control parameter change Unit 41 Step command unit 42 Speed PI control unit 43 Current PI control unit 44 Motor model 45 Mechanical system model 100, 200 Motor control device 101 Machine learning device

Claims (11)

It is a machine learning device that performs machine learning of the motor control device by connecting it to a motor control device that controls a motor that is a permanent magnet synchronous motor by vector control.
A speed response time series data storage unit that stores response time series data of the rotation speed of the motor,
A speed time constant detection unit that detects the time constant of speed control from the response time series data of the rotation speed,
An inertia constant correction unit that corrects the inertia constant of the motor shaft of the motor based on the time constant output by the speed time constant detection unit.
A speed response data generation unit that generates speed response teacher data and speed response test data based on a preset inertia constant and an inertia correction value output by the inertia constant correction unit.
A speed response classification unit that classifies the speed response of the motor based on the speed response teacher data and the speed response test data, and a speed response classification unit.
A machine learning device equipped with. The speed response classification unit classifies the class based on at least one of the value of the time constant at the time of rising, the amount of overshoot, and the value of the steady deviation when the speed command value is given during the synchronous rotation of the motor. The classification is output to the motor control device.
The machine learning device according to claim 1. The inertia constant correction unit identifies the true inertia value of the motor shaft of the motor from the time constant of the speed time constant detection unit, outputs the output to the motor control device, and controls the motor.
The machine learning device according to claim 1 or 2. The velocity response teacher data is a rising time constant, an overshoot amount, and a rising time constant in transient response time series data of an integral system or a first-order lag system generated based on a predetermined velocity response frequency and one or more inertia settings. The classification based on the feature amount vector including at least one of the constant deviation values is output to the motor control device.
The machine learning device according to claim 1. It is a machine learning device that performs machine learning of the motor control device by connecting it to a motor control device that controls a motor that is a permanent magnet synchronous motor by vector control.
A current response time-series data storage unit that stores the response time-series data of the dq coordinate current of the motor,
A current time constant detector that detects the time constant of current control from the time series data of the dq coordinate current, and
A motor constant correction unit that corrects the resistance value of the motor and the q-axis inductance based on the time constant output by the current time constant detection unit.
A current response data generation unit that generates current response teacher data and current response test data based on a preset motor constant setting value and a motor constant correction value output by the motor constant correction unit.
A current response classification unit that classifies the current response of the motor based on the current response teacher data and the current response test data, and a current response classification unit.
A machine learning device equipped with. In the positioning of the rotor of the motor, the current response classification unit determines the time constant of the rise time of the dq coordinate system current with respect to the time series data of the dq coordinate system current when the current command value is given. The classification based on at least one of the overshoot amount and the steady deviation value is output to the motor control device.
The machine learning device according to claim 5. The motor constant correction unit obtains at least one of the true resistance value of the motor, the true inductance value of the d-axis of the motor, and the true inductance value of the q-axis of the motor from the time constant of the current time constant detection unit. Identify and output to the motor control device,
The machine learning device according to claim 5. The current response teacher data is
Rising time constant, overshoot amount, and rise time constant in transient response time series data of integral system or first-order lag system generated based on at least one of resistance set value, d-axis inductance set value, and q-axis inductance set value. The class classification classified based on the feature quantity vector including at least one of the constant deviation values and the predetermined current response frequency are output to the motor control device.
The machine learning device according to claim 5 or 6. The machine learning device according to claim 1, wherein the motor control device controls the motor by sensorless vector control and connects to the motor control device to perform machine learning. A motor control device that controls a motor, which is a permanent magnet synchronous motor, by vector control according to the output from the machine learning device according to any one of claims 1 to 8.
A current detector that detects three-phase currents of U-phase, V-phase, and W-phase in the stationary coordinate system of the motor, and
A rotation position detection unit that detects the actual rotation position of the motor,
A three-phase / dq coordinate conversion unit that converts the three-phase current into a d-axis current and a q-axis current in a synchronous rotating coordinate system.
A speed control unit that calculates at least one command current of the d-axis current or the q-axis current based on the rotation speed calculated from the speed command value of the motor and the actual rotation position.
A current control unit that calculates at least one voltage command on the d-axis or the q-axis based on the command current and the d-axis current or the q-axis current.
A dq coordinate / three-phase conversion unit that converts the voltage command into a three-phase voltage command of the synchronous rotating coordinate system,
Based on the three-phase voltage command, the PWM control unit that controls the pulse width of the switching element provided in the inverter and drives the motor.
An inertia constant changing unit that changes the motor shaft inertia constant of the motor,
A speed control parameter changing unit that changes the parameters of the speed control unit,
A motor constant changing unit that changes the resistance constant, d-axis inductance constant, and q-axis inductance constant of the motor,
A current control parameter changing unit that changes the parameters of the current control unit,
A changeability determination unit that determines whether or not at least one of the speed control parameter change unit, the motor constant change unit, and the current control parameter change unit can be executed.
A motor control device. A motor control device for controlling a motor, which is a permanent magnet synchronous motor, by sensorless vector control according to the output from the machine learning device according to claim 9.
A current detector that detects three-phase currents of U-phase, V-phase, and W-phase in the stationary coordinate system of the motor, and
A three-phase γδ coordinate conversion unit that converts the three-phase current into a γ-axis current and a δ-axis current in an estimated rotating coordinate system.
An induced voltage observer unit that calculates at least one velocity-induced voltage of the γ-axis velocity-induced voltage or the δ-axis velocity-induced voltage based on the γ-axis current and the δ-axis current.
A rotation position estimation unit that calculates the estimated rotation position of the motor based on the velocity-induced voltage, and
A rotation speed estimation unit that calculates the estimated rotation speed of the motor based on the speed-induced voltage, and a rotation speed estimation unit.
An estimated speed control unit that calculates at least one command current of the γ-axis current or the δ-axis current based on the speed command value of the motor and the estimated rotation speed calculated by the rotation speed estimation unit.
A γδ-axis current control unit that calculates at least one estimated voltage command value of the γ-axis or the δ-axis based on the command current and the γδ-axis current.
A γδ coordinate / three-phase conversion unit that converts the estimated voltage command value into a three-phase voltage command value of the estimated rotating coordinate system,
A PWM control unit that controls the pulse width of the switching element provided in the inverter based on the three-phase voltage command value and drives the motor.
An estimated speed control parameter changing unit that changes the parameters of the estimated speed control unit,
The γδ axis current control parameter changing unit that changes the parameters of the γδ axis current control unit,
An inertia constant changing unit that changes the motor shaft inertia constant of the motor,
A motor constant changing unit that changes the resistance constant, d-axis inductance constant, and q-axis inductance constant of the motor,
A changeability determination unit that determines whether or not at least one of the estimated speed control parameter change unit, the motor constant change unit, and the γδ axis current control parameter change unit can be executed.
A motor control device.

Images