Research on Unbalanced Electromagnetic Force Under Static Eccentricity of the Wheel Hub Motor Based on BP Neural Network

Abstract

Aiming at exploring a high-precision unbalanced electromagnetic force model suitable for the dynamic simulation of wheel hub direct-drive electric vehicles, this article establishes the unbalanced electromagnetic force model under static eccentricity of a wheel hub motor by an analytical method and verifies its accuracy by finite element modeling. Then, it optimizes the unbalanced electromagnetic force model based on a BP neural network and couples it with the 1/2 vehicle vertical vibration model to improve its calculation and operation efficiency. Finally, the correctness of the coupling model is further verified by bench experiments. The results show that the analytical model of the unbalanced electromagnetic force is accurate. A BP neural network optimization algorithm can reduce the time of electromagnetic force model simulation for 10 s from 1 h to about 50 s, which greatly improves the calculation efficiency of the electromagnetic force on the basis of ensuring the accuracy of the model, thus providing an unbalanced electromagnetic force model that is more suitable for the dynamic simulation of wheel hub direct-drive electric vehicles, which effectively solves the problem that the traditional electromagnetic force is difficult to couple with the vehicle dynamics model and lays a better foundation for subsequent research on the vertical vibration effect of wheel hub direct-drive electric vehicles.

Keywords

• wheel hub motor
• unbalanced electromagnetic force
• finite element verification
• BP neural network
• bench experiment