CN-121997565-A - Motor performance index prediction method and device

CN121997565ACN 121997565 ACN121997565 ACN 121997565ACN-121997565-A

Abstract

The invention relates to the technical field of motors and provides a motor performance index prediction method and a motor performance index prediction device, wherein the method comprises the steps of obtaining design parameters of a motor, wherein the design parameters of the motor comprise at least one of geometric parameters, material parameters, environment parameters and working condition parameters of the motor; the method comprises the steps of inputting design parameters into a performance index prediction model to obtain the amplitude and the phase of the air gap magnetic flux density harmonic wave of a motor output by the performance index prediction model, and calculating the performance index of the motor based on the amplitude and the phase of the air gap magnetic flux density harmonic wave, wherein the performance index prediction model is obtained by training based on design parameter samples and amplitude labels and phase labels of the air gap magnetic flux density harmonic wave corresponding to the design parameter samples. The invention can accurately predict the motor performance index according to the design parameters of the motor.

Inventors

YAN LINA
LI HUI
YAN YIFAN
REN KUNHUA
QI HONGFENG
LIU BAIBO

Assignees

中车工业研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20251230

Claims (10)

1. The motor performance index prediction method is characterized by comprising the following steps of: Obtaining design parameters of a motor, wherein the design parameters of the motor comprise at least one of geometric parameters, material parameters, environment parameters and working condition parameters of the motor; Inputting the design parameters into a performance index prediction model to obtain the amplitude and the phase of the air gap magnetic flux density harmonic wave of the motor output by the performance index prediction model; calculating a performance index of the motor based on the amplitude and phase of the air gap flux density harmonic; The performance index prediction model is obtained by training based on design parameter samples and amplitude labels and phase labels of air gap magnetic flux density harmonic waves corresponding to the design parameter samples.
2. The method of claim 1, wherein the amplitude and phase of the air gap flux density harmonic comprise the amplitude and phase of a radial flux density harmonic and/or the amplitude and phase of a tangential flux density harmonic; The amplitude and phase of the radial magnetic flux density harmonic wave comprise the amplitude and phase of a fundamental wave order matched with the pole pair number of the motor, the amplitude and phase of a slot harmonic wave order matched with the slot number of the stator, and the amplitude and phase of a combination order of the fundamental wave and the slot harmonic wave, wherein the amplitude and phase of the fundamental wave order are related with the radial magnetic flux density; the amplitude and phase of the tangential magnetic flux density harmonics include the amplitude and phase of the fundamental wave order matching the pole pair number of the motor, the amplitude and phase of the slot harmonic order matching the stator slot number, and the amplitude and phase of the combined order of both the fundamental wave and slot harmonic.
3. The method according to claim 2, wherein the performance index prediction model comprises a first prediction model and/or a second prediction model; The first prediction model is used for predicting the amplitude and the phase of radial magnetic flux density harmonic waves based on the design parameters; the second predictive model is used to predict the magnitude and phase of the tangential magnetic flux density harmonic based on the design parameters.
4. The method according to claim 2, wherein the performance index includes at least one of a no-load back electromotive force waveform, a cogging torque waveform, a core loss, a permanent magnet eddy current loss, an electromagnetic exciting force, and a vibration noise of the motor; calculating a performance index of the motor based on the amplitude and phase of the air gap flux density harmonic, including at least one of the following calculation modes: based on the amplitude and the phase of the radial magnetic flux density harmonic wave, synthesizing a radial air gap magnetic flux density distribution function, calculating the radial air gap magnetic flux density distribution function through integration and deriving to obtain an idle-load back electromotive force waveform and an effective value of the motor; Combining the harmonic characteristics of the radial magnetic flux density and the tangential magnetic flux density, and calculating to obtain a cogging torque waveform and a peak value of the cogging torque waveform of the motor by adopting a Max Wei Zhangliang method; Based on the amplitude and the phase of radial magnetic flux density harmonic waves, calculating to obtain the core loss in the stator and rotor cores by adopting a Stankmez or Bessel function decomposition model; Based on the amplitude and frequency of radial magnetic flux density harmonic wave induced in the permanent magnet area and changing with time, the eddy current loss of the permanent magnet in the permanent magnet can be calculated by combining the conductivity and the segmentation condition of the permanent magnet; And combining the respective amplitude and phase of the radial magnetic flux density harmonic wave and the tangential magnetic flux density harmonic wave, calculating radial electromagnetic force density distribution acting on the tooth part of the stator according to a Maxwell stress formula, and carrying out two-dimensional Fourier decomposition of space and time domain on the radial electromagnetic force density distribution to obtain force wave amplitude values of each order and each frequency, and evaluating electromagnetic vibration and vibration noise of the motor based on the force wave amplitude values.
5. The motor performance index prediction method according to any one of claims 1 to 4, characterized in that the performance index prediction model is trained as follows: acquiring an amplitude label and a phase label of an air gap magnetic flux density harmonic corresponding to a design parameter sample; Inputting the design parameter sample into an initial prediction model to obtain an amplitude predicted value and a phase predicted value of the air gap magnetic flux density harmonic wave output by the initial prediction model; substituting the amplitude predictive value and the phase predictive value, and the corresponding amplitude label and phase label into a loss function, reversely propagating when the loss function is not converged, and adjusting model parameters of an initial predictive model until the loss function is converged or iterated for a preset number of times, so as to obtain a trained performance index predictive model.
6. The method of claim 5, wherein obtaining the design parameter samples and the amplitude and phase labels of the air gap flux density harmonics corresponding to the design parameter samples comprises: sampling the original design data of the motor to obtain design parameter samples of different parameter combinations; performing finite element simulation on the design parameter sample to obtain an air gap magnetic flux density data sample; And performing fast Fourier transform on the air gap magnetic flux density data samples based on an air gap magnetic field adjustment theory to obtain amplitude labels and phase labels of air gap magnetic flux density harmonic waves.
7. The method of claim 6, further comprising, after obtaining the amplitude tag and the phase tag of the air gap flux density harmonic: And for each group of design parameter samples, the first N maximum amplitude labels and the corresponding N phase labels are screened from the amplitude labels and the phase labels of all the corresponding air gap flux density harmonic waves and are used for training a performance index prediction model, wherein N is more than or equal to 2.
8. A motor performance index prediction apparatus, comprising: The design parameter acquisition module is used for acquiring design parameters of the motor, wherein the design parameters of the motor comprise at least one of geometric parameters, material parameters, environment parameters and working condition parameters of the motor; the model prediction module is used for inputting the design parameters into a performance index prediction model to obtain the amplitude and the phase of the air gap magnetic flux density harmonic wave of the motor output by the performance index prediction model; the performance index calculation module is used for calculating the performance index of the motor based on the amplitude and the phase of the air gap magnetic flux density harmonic wave; The performance index prediction model is obtained by training based on design parameter samples and amplitude labels and phase labels of air gap magnetic flux density harmonic waves corresponding to the design parameter samples.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor implements the motor performance index prediction method of any one of claims 1 to 7 when the computer program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the motor performance index prediction method according to any one of claims 1 to 7.

Description

Motor performance index prediction method and device Technical Field The invention relates to the technical field of motors, in particular to a motor performance index prediction method and device. Background In modern industrial fields such as rail transit equipment, low-altitude aircrafts, new energy automobiles, precise transmission equipment, intelligent household appliances and the like, a motor is used as a core power component, and the running stability, energy efficiency and noise vibration characteristics of the motor directly influence the competitiveness of the whole machine product. With the continuous development of the motor application field to high efficiency, small size and precision, the requirements on core performances such as no-load back electromotive force and cogging torque of the motor are increasingly increased, and especially in high-rotation speed and high-precision control occasions, small deviation of performance indexes can cause abnormal system operation. Accurate predictions of core performance metrics must be achieved during the design phase. The traditional performance index prediction method has obvious defects, although the finite element simulation can reflect electromagnetic field distribution in a finer way, modeling is complex, calculation is time-consuming, and the requirements of multi-scheme rapid comparison and motor design parameter optimization are difficult to meet. The traditional machine learning model such as linear regression and the like tries to establish the mapping between design parameters (such as geometric parameters and material parameters) and performance indexes of the motor, but fails to deeply characterize harmonic characteristics of an air-gap magnetic field, and the amplitude and phase coupling effect of multi-order harmonic waves are characterized insufficiently, so that the model generalization capability is weak, and the prediction error in an unknown design parameter interval is larger. Disclosure of Invention The invention provides a motor performance index prediction method and device, which are used for solving the problems of weak generalization capability and inaccurate prediction of a motor performance index prediction model in the prior art. The invention provides a motor performance index prediction method, which comprises the following steps: Obtaining design parameters of a motor, wherein the design parameters of the motor comprise at least one of geometric parameters, material parameters, environment parameters and working condition parameters of the motor; Inputting the design parameters into a performance index prediction model to obtain the amplitude and the phase of the air gap magnetic flux density harmonic wave of the motor output by the performance index prediction model; calculating a performance index of the motor based on the amplitude and phase of the air gap flux density harmonic; The performance index prediction model is obtained by training based on design parameter samples and amplitude labels and phase labels of air gap magnetic flux density harmonic waves corresponding to the design parameter samples. According to the motor performance index prediction method provided by the invention, the amplitude and the phase of the air gap magnetic flux density harmonic comprise the amplitude and the phase of the radial magnetic flux density harmonic and/or the amplitude and the phase of the tangential magnetic flux density harmonic; The amplitude and phase of the radial magnetic flux density harmonic wave comprise the amplitude and phase of a fundamental wave order matched with the pole pair number of the motor, the amplitude and phase of a slot harmonic wave order matched with the slot number of the stator, and the amplitude and phase of a combination order of the fundamental wave and the slot harmonic wave, wherein the amplitude and phase of the fundamental wave order are related with the radial magnetic flux density; the amplitude and phase of the tangential magnetic flux density harmonics include the amplitude and phase of the fundamental wave order matching the pole pair number of the motor, the amplitude and phase of the slot harmonic order matching the stator slot number, and the amplitude and phase of the combined order of both the fundamental wave and slot harmonic. According to the motor performance index prediction method provided by the invention, the performance index prediction model comprises a first prediction model and/or a second prediction model; The first prediction model is used for predicting the amplitude and the phase of radial magnetic flux density harmonic waves based on the design parameters; the second predictive model is used to predict the magnitude and phase of the tangential magnetic flux density harmonic based on the design parameters. The motor performance index prediction method comprises at least one of no-load back electromotive force waveform, cogging torque waveform, core loss, perma