Optimization modeling method and system for electro-magnetic doubly salient reluctance motor

Abstract

The invention relates to an optimization modeling method of an electro-magnetic doubly salient reluctance motor. The method comprises the following steps: selecting a plurality of optimization targets according to an application scene of the motor; selecting the first n structural parameters with the maximum numerical values from the structural parameters of the motor body by calculating the comprehensive sensitivity index; each parameter structure is preset to have M selection levels, and Mn groups of parameter modeling samples can be obtained; normalizing all parameter modeling samples, and randomly dividing the parameter modeling samples into a training set and a test set according to a set proportion; and carrying out machine learning on the training set and a set optimization target by utilizing a KNN algorithm, establishing a motor optimization model, and checking the accuracy of the established motor optimization model through the test set. Compared with the prior art, the method has the advantages that the comprehensive sensitivity index is introduced to reduce the structural parameters of the motor body, the parameter modeling sample space is greatly reduced, the data required by the sample is effectively reduced, the modeling speed is increased, and the requirement of the industry for rapid optimization of the motor is met.

Description

technical field [0001] The invention relates to the field of reluctance motors, in particular to an optimization modeling method and system for an electrically excited double salient pole reluctance motor. Background technique [0002] Compared with the doubly salient pole motor using permanent magnet excitation and hybrid excitation, the electric excitation motor can easily adjust the excitation current to change the output voltage, and there is no need to worry about the demagnetization of the permanent magnet caused by the motor working in a high temperature and harsh environment . Therefore, the reluctance represented by electrically excited double convex poles has broad application prospects in aerospace, wind power generation, and new energy vehicles. However, electric excitation motors also have their own inherent defects. Compared with permanent magnet excitation and hybrid excitation, the power density and output efficiency of electric excitation motors still have ...

AI Technical Summary

Problems solved by technology

There are some problems in the above two methods. Using finite element simulation software to build model data has higher accuracy, but finite element simulation is time-consuming. When there is a lot of modeling data, the simulation workload is particularly large.

The finite element simulation modeling is only to compare and analyze the discrete motor mechanism parameters, the selected samples are relatively small, and the model is not established for all parameters within the allowable value range, which cannot meet the real-time requirements of the subsequent optimization process; although the analytical method It meets the subsequent real-time requirements, but it is solved under the assumption of ideal conditions, so the accuracy is not high, which leads to the result that may not be the global optimal structure in the subsequent optimization process

Images