JP2012210067A - Pulsation suppression device of motor and pulsation suppression method of motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a pulsation suppression device of a motor capable of estimating a system identification model error.SOLUTION: The pulsation suppression device includes a periodic disturbance observer PDO which estimates a disturbance torque by multiplying an axial torque Tof a motor (actual plant P) driven by vector control of an inverter with an inverse function of a frequency transmission function from a torque ripple compensation current command value to an axial torque detection value, an error estimator 100 which acquires a system identification model error ΔPby comparing an axial torque Tin an initial state just before starting of torque ripple suppression control with an axial torque estimation value T^estimated from a system identification model P^by assuming a disturbance estimation value dI^estimated by the PDO as a disturbance in a final condition where a torque ripple is sufficiently canceled after starting of the torque ripple suppression control, and a memory 110 in which the error ΔPis stored. The system identification model P^of the PDO is corrected by the error ΔPin the memory 110.

Description

  The present invention relates to a control device and a control method for automatically suppressing torque pulsation (hereinafter also referred to as torque ripple) of an electric motor or the like.

  Conventionally, as a pulsation suppressing device for an electric motor, for example, as disclosed in Patent Document 1, the pulsation of the electric motor is extracted by Fourier transform for an arbitrary order component, and learning control is performed so that the Fourier coefficient becomes 0, There is known a device for adding a pulsation compensation signal (compensation current) obtained by the control to an inverter current command.

  In Non-Patent Document 1, the system identification from the torque ripple compensation command to the shaft torque detection value is performed in advance, the frequency component of the torque pulsation of the shaft torque detection value is extracted, and this is multiplied by the inverse system of the system identification model. Thus, it is disclosed that a disturbance torque is estimated and a torque ripple compensation command is generated based on a difference between the estimated value of the disturbance torque and a command value (usually 0 value).

International Publication WO2010 / 024195A1

Y. Tadano, et al. “Periodic Learning Suppression Control of Torque Ripple Optimized System Identification for Permanent Magnet Synchronous Motors, IEEE IPEC. 1363-1370 (2010)

  The techniques disclosed in Patent Document 1 and Non-Patent Document 1 will be described with reference to FIGS. FIG. 1 shows the configuration of a control target device for torque ripple suppression control.

  In FIG. 1, a motor 2 (for example, PMSM (Permanent Magnet Synchronous Motors) driven by an inverter 1 and a source of torque pulsation) and a load device 3 are coupled by a shaft 4.

  The shaft torque of the shaft 4 is measured by a shaft torque meter 5 and input as a torque detection value to the torque ripple suppression device 6, and the rotor position of the motor 2 is measured by a rotational position sensor 7 such as a rotary encoder to detect the phase. The value is input to the torque ripple suppression device 6 as a value.

  The torque ripple suppression device 6 includes a torque pulsation frequency component extraction unit that extracts a torque pulsation frequency component by Fourier transform based on the input torque detection value and phase detection value, and an inverse of the system identification model for the extracted component. A torque ripple suppression control unit having a periodic disturbance observer that multiplies the system to estimate disturbance torque, and a compensation current generation unit that generates a compensation current from the estimated disturbance torque value and a compensation current command value and applies the compensation current to the inverter 1. ing.

  In the example of FIG. 1, in consideration of the current vector control by the inverter 1, the d-axis and q-axis current command values (compensation current) on the rotation coordinates (orthogonal dq axes) synchronized with the rotation of the motor 2 are given. Yes.

  In FIG. 1, torque pulsation is detected by feedback from the shaft torque meter 5, but this form is only an example, frame vibration detection by an acceleration sensor or the like installed on the frame, and rotational speed / position fluctuation by a rotational position sensor or the like. Detection or replacement with current pulsation detection by current sensor, motor shaft pulsation suppression, motor frame pulsation suppression, motor speed pulsation suppression or rotational position pulsation suppression, motor current pulsation control It is.

  For the inverter 1, for example, a general control device that controls motor current (torque) by current vector control that realizes maximum torque / current control may be used.

  The torque ripple suppression device 6 extracts a torque pulsation frequency component and generates a compensation current for suppressing it for each frequency component.

  FIG. 2 shows a basic configuration of the torque ripple suppression control system in FIG. 1, and FIG. 3 shows a control block representing only the torque ripple frequency component of the periodic disturbance observer in FIG.

  Definitions of symbols in FIGS. 2 and 3 are as follows.

T ^* : Torque command value, T _det : Torque detection value,
T _An : n-th order torque pulsation extraction component (cosine coefficient),
T _Bn : n-th order torque pulsation extraction component (sine coefficient),
ω: rotational speed detection value, θ: rotational phase detection value,
id, iq: d, q-axis current detection value, id ^* , iq ^* : d, q-axis current command value,
iq ₀ ^* : q-axis current command value (before compensation current superposition),
iqc ^* : Torque ripple compensation current command value I _An : n-order compensation current cosine coefficient, I _Bn : n-order compensation current sine coefficient,
iu, iv, iw: u, v, w-phase current abz: rotation sensor signal dI ^* _An, dI ^* _Bn : n-order compensation current command value,
dI ^ _An, dI ^ _Bn : n-th order disturbance current estimated value,
dI _n : n-order disturbance current The subscript n means the n-order frequency component.

  2 and 3, the same parts as those in FIG. 1 are denoted by the same reference numerals, 10 is an actual system, 11 is a torque to id, iq conversion unit, 12 is a current vector control unit, 13 is a current sensor, 14 Is a coordinate conversion unit, 15 is a rotation phase / speed detection unit, 61 is a torque pulsation frequency component extraction unit, 62 is a torque ripple suppression control unit, 63 is a compensation current generation unit, and 621 is a periodic disturbance observer (Periodic Disturbance Observer). ing.

2 and 3, the torque-to-id, iq converter 11 converts the torque command value T ^* into a desired d-axis current command value id ^* and q-axis current command value iq ₀ ^* . In general, conversion formulas and tables that realize maximum torque / current control are used. The current vector control unit 12 performs current control on each axis in the general orthogonal rotating coordinate system d-axis q-axis. The coordinate conversion unit 14 converts the three-phase alternating currents iu, iv, iw into currents id, iq in a dq axis orthogonal rotation coordinate system synchronized with the motor rotation coordinates. The rotation phase / speed detection unit 15 converts information from the rotation position sensor 7 such as an encoder into information on the rotation phase and the rotation speed.

It is known that the torque ripple of the motor is a disturbance that periodically occurs according to the rotational phase θ [rad]. Therefore, a means for extracting a torque ripple frequency component in synchronism with the motor rotational speed ω [rad / s], that is, a torque pulsation frequency component extraction unit 61 is used, and a cosine coefficient T of an arbitrary order n (an integer multiple of the electrical rotational frequency). _An [Nm] and a sine coefficient T _Bn [Nm] are converted. The strict measurement means of the frequency component includes Fourier transform, but in this example, importance is placed on the ease of calculation, and the n-th order based on the rotational phase θ as the shaft torque detection value T _det [Nm] by the shaft torque meter 5. multiplied by the cosine wave-sine wave, by performing a low-pass filter G _F respectively, it is carried out an approximate Fourier transform.

Next, before performing control, the transfer characteristic of the actual system 10 from the torque ripple compensation current command value iqc ^* to the shaft torque detection value T _det is identified in advance. Since the system identification method is a general technique, an arbitrary method may be used.

  Here, the frequency transfer function of the real system 10 is defined as a complex vector P (jω) with respect to the rotational speed ω as shown in equation (1).

P (jω) = P _A (jω) + jP _B (jω) (1)
* P _A : Real part of real system, P _B : Imaginary part of real system.

Since the suppression control system constructs the control system with torque ripple synchronous coordinates, only the frequency transfer characteristic of the arbitrary n-th order component is extracted from the expression (1), and the actual system of the torque ripple synchronous coordinates is represented by the expression (2). That is, the amplitude / phase characteristics of the system of an arbitrary n-order component can be expressed by a simple one-dimensional complex vector P _n . The real part / imaginary part axes are defined with reference to the rotational phase, and the cosine coefficient corresponds to the real part component and the sine coefficient corresponds to the imaginary part component.

P _n = P _An + jP _Bn (2)
* P _An : real system n-order component real part, P _Bn : real system n-order component imaginary part.
Similarly, the system identification result in the suppression control system is defined as a one-dimensional complex vector as shown in equation (3).

P ^ _n = P ^ _An + jP ^ _Bn (3)
* P ^ _An : n-order component real part of identification result, P ^ _Bn : n-order component imaginary part of identification result.

  For example, when the system identification result of 1 to 1000 Hz is expressed as a complex vector every 1 Hz, a table composed of 1000 one-dimensional complex vector elements can be constructed. It is also possible to express the identification result by an approximate expression. Therefore, even in a complex system, the system model can always be expressed by a simple one-dimensional complex vector.

Disturbance torque is estimated in the form of a complex vector by multiplying the extracted and converted T _An and T _Bn by the inverse system of identification model (1 / (P _An + jP _Bn )). These processes are performed for each order component, and the compensation current generator 63 adds the deviation between the estimated disturbance torque values (dI ^ _An, dI ^ _Bn ) and the compensation current command values (dI ^* _An, dI ^* _Bn ) to the adder 17. _a, 17 to generate a torque ripple compensation current command value iqc ^* cosine coefficients of n order frequency components of the ^{_{[a] iqc n * [a}} ] I an * [a] and sine coefficients I _Bn ^* [a] and determined by _B. The conversion to the n-order compensation current command value iqcn ^* is calculated using the same rotational phase θ as in the torque ripple synchronous coordinate conversion.

Incidentally, the I _An ^*, I _Bn ^*, through the low-pass filter G _F, are used by the disturbance observer 621 internally as component that does not pass through the actual system 10.

The converter 11 converts the torque command T ^* [NM] into dq-axis current command values id ^* and iq ₀ ^* that realize the maximum torque / current control, and a combined value of the generated respective compensation current command values. Normal vector control is performed by superimposing iqc ^* on iq ₀ ^* .

  Basically, the contents of calculation processing performed by the torque ripple suppression device 6 are shaft torque pulsation component extraction, torque ripple suppression, and compensation current signal generation, and other processing is performed by a general inverter.

In this example, the compensation current command value iqc ^* is superimposed on the q-axis current command value iq ₀ ^* , but it is replaced with the d-axis current command value, the current command values for both the d-axis and the q-axis, the torque command value, and the like. Is also possible.

  In the high frequency band, the feedback control response decreases due to the influence of the computation dead time. In general, in the method of performing feedback suppression control in the entire frequency band, the disturbance suppression performance in the high frequency band is reduced, and it is difficult to generate a desired compensation current. On the other hand, the control method handled in FIGS. 1 to 3 generates a sine wave / cosine wave having the same frequency as the periodic disturbance, and then adjusts the sine / cosine coefficient (equivalent to amplitude / phase). Therefore, this method is easier to deal with high frequency disturbances than a controller that collectively generates compensation currents in all frequency bands, and can be expected to improve periodic disturbance suppression performance. Further, if a suppression controller (torque ripple suppression device 6) is configured in parallel for each target order, it is possible to simultaneously cope with torque ripple suppression of multiple orders.

  Regarding the control systems shown in FIGS. 2 and 3, what constitutes the basis of control and affects the control performance is the accuracy with respect to the true value of the system identification model. In order to improve torque ripple suppression capability, more accurate system identification is required.

  However, it is difficult to obtain the identification model with high accuracy, and the system is expected to fluctuate due to secular change. The error from the true value may increase the convergence time until the suppression is completed, or in the worst case, the suppression control itself may increase the torque ripple due to the phase error, which may make the control unstable. For this reason, improvement in robustness against identification model errors is required.

  The present invention solves the above-described problems, and an object thereof is to provide a pulsation suppressing device and a pulsation suppressing method for an electric motor capable of estimating a system identification model error.

  The pulsation suppressing device for an electric motor according to claim 1 for solving the above-described problem is a d, q-axis current command value obtained by converting a torque command value of the motor into a d, q-axis current component of a rotating coordinate system in vector control. And a torque pulsation compensation current command value identified as a system identification model in an inverter that performs current control of the motor from the output current detection value of the inverter and a torque pulsation component extracted by Fourier transform of the torque pulsation of the motor A periodic disturbance observer that estimates a periodic disturbance torque by multiplying an inverse function of a frequency transfer function from the motor to the detected shaft torque of the motor, and a disturbance torque estimated value and a compensation current command value estimated by the periodic disturbance observer Torque pulsation compensation control means for adding the torque pulsation compensation current command value generated from the above to the d and q axis current command values, and the transfer function is identified. The shaft torque detection value in the initial state immediately before the start of the torque pulsation suppression control by the torque pulsation suppression control means in the stem, after the torque pulsation suppression control by the torque pulsation suppression control means has started, and the torque pulsation has been canceled An error estimation means for obtaining a system identification model error by regarding a disturbance torque estimated value estimated by the periodic disturbance observer as a disturbance in a final state and comparing the estimated value with a shaft torque estimated value from the system identification model; It is characterized by having prepared.

  According to a fifth aspect of the present invention, there is provided a method for suppressing pulsation of an electric motor, wherein d and q-axis current command values obtained by converting a torque command value of the motive force into d and q-axis current components of a rotating coordinate system in vector control and output current detection of an inverter Inverter for controlling current of the motor from the value, and torque pulsation component extracted by Fourier transform of the torque pulsation of the motor, identified as a system identification model, shaft torque detection of the motor from the torque pulsation compensation current command value A periodic turbulence observer that estimates the periodic disturbance torque by multiplying the inverse function of the frequency transfer function up to the value, and the torque pulsation compensation current generated from the estimated disturbance torque estimated by the periodic disturbance observer and the compensation current command value A method for suppressing pulsation of an electric motor in an apparatus comprising torque pulsation suppression control means for adding a command value to the d and q-axis current command values. In the actual system in which the transfer function is identified, the estimation unit detects the axial torque detected value in the initial state immediately before the torque pulsation suppression control by the torque pulsation suppression control starts, and the torque pulsation suppression control by the torque pulsation suppression control unit After the start and in the final state where the torque pulsation is canceled, the disturbance torque estimated value estimated by the periodic disturbance observer is regarded as a disturbance, and compared with the shaft torque estimated value estimated from the system identification model, An error estimation step for obtaining a system identification model error is provided.

  According to the above configuration, the system identification model error can be estimated.

  According to a second aspect of the present invention, there is provided the motor pulsation suppressing device according to the first aspect, wherein the system identification model error obtained by the error estimating means is stored as table data using the rotation speed and torque of the motor as reference values. Storing means for correcting the system identification model using a system identification model error stored in the storage means.

  According to a sixth aspect of the present invention, there is provided a method for suppressing pulsation of an electric motor according to the fifth aspect, wherein the error estimation means uses the rotation speed and torque of the electric motor as reference values for the system identification model error obtained in the error estimation step. A storage step of storing the data in the storage unit as table data is provided, and the system identification model is corrected using a system identification model error stored in the storage unit.

  According to the said structure, a system identification model is correct | amended and optimal suppression control can be performed with an exact identification model.

  According to a third aspect of the present invention, there is provided the pulsation suppressing device for an electric motor according to the second aspect, wherein the error estimating means is provided for a plurality of orders of torque pulsation frequency components to obtain a system identification model error in the torque pulsation frequency component of each order. The system identification model is corrected.

  According to a seventh aspect of the present invention, there is provided the method for suppressing pulsation of an electric motor according to the sixth aspect, wherein the error estimating means is provided for a plurality of orders of torque pulsation frequency components, and the storing step is a system for torque pulsation frequency components of the respective orders. The identification model error is stored in the storage means.

  According to the above configuration, even when there are a plurality of torque ripple frequency components for which suppression and system identification model errors are to be estimated, torque ripples of a plurality of orders can be suppressed.

  According to a fourth aspect of the present invention, there is provided the motor pulsation suppressing device according to the second or third aspect, wherein the error estimating means is a torque generated by the torque pulsation control means at a frequency / torque operating point within an operating range of the electric motor. The shaft torque detection value in the initial state immediately before the start of the pulsation suppression control is recorded in the storage means as an initial value, and after the torque pulsation suppression control is started by the torque pulsation suppression control means at the operating point, and the torque pulsation is The estimated shaft torque estimated at the time of the canceled final state is recorded as the final value in the storage unit, the initial value is compared with the final value, a system identification model error is calculated, and the error calculation result is stored in the storage unit. After saving the system identification model error, change the operating point without performing the torque pulsation suppression control, and at the operating point after the change The initial value, the recording of the final value, is characterized by executing the storage of calculation and said error calculation result of system identification model error.

  The motor pulsation suppression method according to an eighth aspect of the present invention is the motor pulsation suppression method according to the sixth or seventh aspect, wherein the error estimating means is a torque pulsation by the torque pulsation control means at a frequency / torque operating point within an operating range of the electric motor. An initial value recording step of recording in the storage means the detected shaft torque value in the initial state immediately before the start of the suppression control, and after the start of the torque pulsation suppression control by the torque pulsation suppression control means at the operating point; And a final value recording step for recording in the storage means the estimated shaft torque value estimated in the final state where the torque pulsation was canceled, and an error for calculating the system identification model error by comparing the initial value and the final value. A calculation step, a data storage step for storing the error calculation result in the storage means, and after storing the system identification model error The operating point is changed without performing the torque pulsation suppression control, and the initial value recording step, the final value recording step, the error calculation step, and the data storage step are executed at the operating point after the change. Yes.

  According to the configuration described above, accurate system identification can be performed even when system identification cannot be performed thoroughly.

(1) According to the first to eighth aspects of the invention, the system identification model error can be estimated.
(2) According to the invention described in claims 2 and 6, the system identification model can be corrected based on the estimated system identification model error, and optimal suppression control can be performed with the accurate identification model.
(3) According to the inventions of claims 3 and 7, even when there are a plurality of torque ripple frequency components for which suppression and system identification model errors are to be estimated, torque ripples of a plurality of orders can be suppressed. .
(4) According to the inventions described in claims 4 and 8, accurate system identification can be performed even when system identification cannot be performed thoroughly. In addition, when multiple torque ripple frequency components for which suppression and system identification model errors are to be estimated exist simultaneously, it is possible to record for multiple frequencies at one operating point, reducing the time to complete the entire measurement. can do.

The block diagram of the control object apparatus to which this invention is applied. The basic block diagram of the torque ripple suppression control system in the apparatus of FIG. FIG. 3 is a control block diagram showing only a torque ripple frequency component of a periodic disturbance observer in the control system of FIG. 2. 1 is a control block diagram showing a first embodiment of the present invention. The control block diagram which shows Example 2 of this invention. The control block diagram which shows Example 3 of this invention. The flowchart which shows the process sequence of Example 4 of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to the following embodiments. In the present embodiment, for example, in the pulsation suppressing device of FIGS. 1 to 3, the time immediately before the start of the suppression control is defined as the initial state, and the state in which the torque ripple is sufficiently canceled after the start of the suppression control is defined as the final state. The identification model is estimated by comparison using error estimation means.

FIG. 4 is a diagram in which the error estimator of the present invention is added to the configuration of FIG. 3 and only frequency components are simplified. In FIG. 4, _Pn is a real plant (real system 10), PDO is a periodic disturbance observer (621), P ^ _n is an identification model (system identification model), and 100 is an error estimator as error estimation means. Yes.

T _n is a shaft torque (detected value) and corresponds to the n-th order torque pulsation extraction components T _An and T _Bn of FIG. dI ^ _n is a disturbance estimated value estimated by the periodic disturbance observer PDO, and corresponds to the nth-order disturbance current estimated values dI ^ _{An and} dI ^ _Bn in FIG.

This dI ^ _n is deviated from the disturbance command value I ^* _{n in} the adder 17 (corresponding to the adders 17 _A and 17 _B in FIG. 3) via the switch sw which is turned on when the suppression control is started.

Differential output of the adder 17 is added to the periodic disturbance component (n Tsugigairan current) dI _n in the adder 18.

T ^ _n is a shaft torque estimated value estimated from the identification model P ^ _n when the disturbance estimated value dI ^ _n is regarded as a disturbance.

The error estimator 100 is the shaft torque T _{n in} the initial state immediately before the start of the torque ripple suppression control (when the switch sw is off), after the start of the torque ripple suppression control (after the switch sw is turned on), and the torque ripple is sufficiently canceled out. Compared with the estimated shaft torque value T ^ n in the final state, an error [Delta] _Pn of the system identification model is output.

In the apparatus configured as described above, the elapsed time from the start of the suppression control is denoted by t, and the actual plant P _n , periodic disturbance component dI _n , and shaft torque T _{n in} the initial state [t = t ₀ ] The relationship is expressed by equation (4). At this time, it is assumed that the time from the initial state to the final state of the suppression control is relatively short, and no time change appears in the plant. Note that the value after the comma of the subscript indicates the value at time t.

dI _{n, t0} · P _n = T _{n, t0} (4)
In the final state [t = t _n ], the shaft torque when the disturbance estimated value dI ^ _{n, tn} is determined as a disturbance by the identification model P ^ n is determined as the shaft torque estimated value T ^ _{n, tn} from the equation (5). Can be estimated.

dI ^ _{n, tn} · P ^ _n = T ^ _{n, tn} (5)
Since there is no change in the disturbance before and after the suppression control at times t ₀ and t _n and the period disturbance is sufficiently suppressed, the disturbance and the disturbance estimated value (compensation command) are the same. It holds.

dI _{n, t0} = dI ^ _{n, tn} (6)
Next, the identification model is expressed by equation (7) as the sum of the true value and the error ΔP _n .

P ^ _n = _Pn + [Delta] _Pn (7)
The expression (7) is substituted into the expression (5) and expressed by the expression (8).

dI ^ _{n, tn.} ( _Pn + [Delta] _Pn ) = T ^ _{n, tn} (8)
Finally, equation (8) is transformed and equation (9) is obtained from the conditions of equations (4) and (6).

From the equation (9), the state model T _{n, t0} before and after the suppression that can be observed and T ^ _{n, tn} are compared by the error estimator 100 to estimate the identification model error, and the correct identification model can be learned.

  In the first embodiment, it is shown that an error of the system identification model is derived by performing suppression control once at a certain rotation speed / torque command. In the second embodiment, the result of error estimation in FIG. 4 is stored in the memory 110 as shown in FIG. 5, the same parts as those in FIG. 4 are denoted by the same reference numerals, and the operation of the error estimator 100 is the same as that in FIG.

The memory 110 uses table data with the rotation speed (ω) / torque (T) as reference values, and refers to this table again at the same rotation speed / torque command. The recorded system identification error is output from the table, and the system identification model P ^ _n (periodic disturbance observer PDO) inside the suppression control is corrected.

  As a result, at the speed / torque point once learned, it is possible to perform optimum suppression control with an accurate identification model thereafter.

In the second embodiment, it is shown that the identification model error can be estimated and corrected for a specific frequency component. In the third embodiment, as shown in FIG. 6, the N-number error estimator with PDO120P ₁ ~120P _N of (order n pieces of the torque ripple frequency component) is provided, the identification model for each order to suppress the control system of FIG. 5 It was configured to perform error estimation in parallel and simultaneously.

The PDOs 120P _{1 to} 120P _N with error estimators are provided with the functions of the PDO of FIG. 5, the error estimator 100 and the memory 110, and an n-order disturbance command value (n-order compensation current command value) I ^The estimated model error is estimated for ^* _{1 to} I ^* _N , and disturbance estimated values dI ^ _{1 to} dI ^ _N obtained as a result of correcting the system identification model based on the error are output.

  According to the above configuration, it is possible to cope with a case where a plurality of torque ripple frequency components for which suppression and system identification model errors are to be estimated exist simultaneously.

  In the fourth embodiment, the identification model error estimation result stored in the table data (stored in the memory 110) in the second and third embodiments is stored in the entire operating range of the device or a certain amount of operating points (the number of rotations).・ Torque), the operation point was changed and repeated.

  Along with the movement of the operating point, the torque ripple suppression control is turned off during the movement of the operating point, and the suppression control is turned on when the operation is stable after the movement, and the initial value and the final value are updated. Complete table data with torque as reference value. The above processing sequence is shown in FIG.

  In FIG. 7, first in step S1, the torque ripple suppression control is turned off (the switch sw in FIGS. 5 and 6 is turned off), and the operating point is moved.

If then the movement of the operating point in step S2 is determined to be completed, in step S3, it is recorded in the memory 110 of the axial torque T _n in the initial state immediately before the torque ripple suppression control starts as an initial value.

  Next, in step S4, the switch sw in FIGS. 5 and 6 is turned on to perform torque ripple suppression control.

Next, when it is determined in step S5 that the suppression has been completed (torque ripple has been sufficiently canceled), in step S6, the estimated shaft torque T ^ _n estimated in the final state is recorded in the memory 110 as the final value. To do.

  Next, in step S7, the switch sw in FIGS. 5 and 6 is turned off to turn off torque ripple suppression control.

In step S8, by comparing the initial value T _n and a final value T ^ _n by error estimator 100 calculates the system identification error [Delta] P _n, is stored in memory 110 in step S9.

  In step S10, it is determined whether or not the processing has been completed for all operating points (all ranges in which the device operates). If the processing has not been completed, the processing in steps S1 to S9 is repeated, and the processing is completed at all operating points. If so, the process ends.

  In addition, when processing is performed at a fixed amount of operating points instead of all the operating points, the shortcomings are interpolated by an arbitrary method. It is also possible to formulate the identification model from the obtained data.

  Even when the system identification cannot be carried out in detail by the above method, accurate system identification can be performed by using the torque ripple suppression control system.

  If the fourth embodiment is applied to the third embodiment, it is possible to record a plurality of frequencies at one operating point, so that it is possible to shorten the time until the entire measurement is completed.

  The present invention is used, for example, for all variable speed devices in which torque pulsation is a problem, for example, suppression of shaft torque resonance of a dynamometer system, suppression of vibration of a motor housing (related to riding comfort such as an electric vehicle and an elevator). be able to.

DESCRIPTION OF SYMBOLS 1 ... Inverter 2 ... Motor 3 ... Load apparatus 4 ... Shaft 5 ... Shaft torque meter 6 ... Torque ripple suppression device 7 ... Rotation position sensor 10 ... Actual system 11 ... Conversion part 12 ... Current vector control part 13 ... Current sensor 14 ... Coordinate conversion part 15 ... rotational phase / speed detector 17,17 _A, 17 _B, 18 ... adder 61 ... torque ripple frequency component extraction unit 62 ... torque ripple suppression control unit 63 ... compensation current generator 100 ... error estimator 110 ... memory 120P _{1 to} 120P _N PDO with error estimator
621: Periodic disturbance observer

Claims (8)

An inverter that controls the current of the motor from the d and q axis current command values obtained by converting the torque command value of the motor into d and q axis current components of the rotating coordinate system in vector control, and the output current detection value of the inverter;
The torque pulsation component extracted by Fourier transform of the torque pulsation of the electric motor is multiplied by the inverse function of the frequency transfer function from the torque pulsation compensation current command value to the detected shaft torque value of the electric motor identified as a system identification model. A periodic disturbance observer for estimating the periodic disturbance torque, and the torque pulsation compensation current command value generated from the disturbance torque estimated value estimated by the periodic disturbance observer and the compensation current command value is used as the d and q axis current command values. Torque pulsation suppression control means to be applied;
In the actual system in which the transfer function is identified, the shaft torque detection value in the initial state immediately before the start of the torque pulsation suppression control by the torque pulsation suppression control unit, and after the start of the torque pulsation suppression control by the torque pulsation suppression control unit In the final state where torque pulsation is canceled, the disturbance torque estimated value estimated by the periodic disturbance observer is regarded as a disturbance, and compared with the shaft torque estimated value estimated from the system identification model. Error estimation means for obtaining
An apparatus for suppressing pulsation of an electric motor, comprising: A storage means for storing the system identification model error obtained by the error estimation means as table data with the rotation speed and torque of the motor as reference values,
The pulsation suppressing device for an electric motor according to claim 1, wherein the system identification model is corrected using a system identification model error stored in the storage unit. 3. The system identification model is corrected according to claim 2, wherein the error estimation means is provided for a plurality of orders of torque pulsation frequency components, a system identification model error in each order of torque pulsation frequency components is obtained, and the system identification model is corrected. Electric motor pulsation suppression device. The error estimation means includes
The shaft torque detection value at the initial state immediately before the start of torque pulsation suppression control by the torque pulsation control means at the operating point of frequency / torque within the operating range of the motor is recorded in the storage means as an initial value,
A shaft torque estimated value estimated in the final state after the torque pulsation suppression control by the torque pulsation suppression control unit at the operating point is started and recorded in the final state where the torque pulsation is canceled is recorded as the final value in the storage unit.
A system identification model error is calculated by comparing the initial value and the final value, and the error calculation result is stored in the storage means,
After storing the system identification model error, the operating point is changed in a state where the torque pulsation suppression control is not performed, and at the changed operating point, the initial value, the final value are recorded, the system identification model error is calculated, and 4. The pulsation suppressing device for an electric motor according to claim 2, wherein the error calculation result is stored. An inverter that controls the current of the motor from the d and q axis current command values obtained by converting the torque command value of the motor into d and q axis current components of the rotating coordinate system in vector control, and the output current detection value of the inverter;
The torque pulsation component extracted by Fourier transform of the torque pulsation of the electric motor is multiplied by the inverse function of the frequency transfer function from the torque pulsation compensation current command value to the detected shaft torque value of the electric motor identified as a system identification model. A periodic disturbance observer for estimating the periodic disturbance torque, and the torque pulsation compensation current command value generated from the disturbance torque estimated value estimated by the periodic disturbance observer and the compensation current command value is used as the d and q axis current command values. A method for suppressing pulsation of an electric motor in a device including torque pulsation suppression control means to be applied,
In the actual system in which the transfer function is identified, the error estimation means is a shaft torque detection value in an initial state immediately before the start of the torque pulsation suppression control by the torque pulsation suppression control means, and torque pulsation suppression by the torque pulsation suppression control means After the start of control and in the final state where the torque pulsation is canceled, the disturbance torque estimated value estimated by the periodic disturbance observer is regarded as a disturbance, and compared with the shaft torque estimated value estimated from the system identification model. And an error estimation step for obtaining a system identification model error. The error estimation means comprises a storage step of storing the system identification model error obtained in the error estimation step in the storage means as table data with the rotation speed and torque of the motor as reference values,
6. The motor pulsation suppression method according to claim 5, wherein the system identification model is corrected using a system identification model error stored in the storage means. The error estimation means is provided for a plurality of orders of torque pulsation frequency components, and the storing step stores system identification model errors in the torque pulsation frequency components of the respective orders in the storage means. The pulsation suppression method of the electric motor described. The error estimation means is
An initial value recording step of recording in the storage means as an initial value a shaft torque detection value in an initial state immediately before starting torque pulsation suppression control by the torque pulsation control means at an operating point of frequency / torque within the operating range of the electric motor When,
A final value record for recording in the storage means the final estimated value of the shaft torque after the start of torque pulsation suppression control by the torque pulsation suppression control means at the operating point and in the final state where the torque pulsation is canceled Steps,
An error calculating step of calculating a system identification model error by comparing the initial value and the final value;
A data storage step of storing the error calculation result in the storage means;
After the system identification model error is stored, the operating point is changed in a state where the torque pulsation suppression control is not performed, and the initial value recording step, the final value recording step, the error calculating step, and data are changed at the operating point after the change. The method for suppressing pulsation of an electric motor according to claim 6 or 7, wherein a storing step is executed.

Images