关键词: 无刷直流电机, 河马优化算法, BP 神经网络, 无感控制

Abstract: To address the problem in traditional sensorless control of brushless direct current motor ( BLDCM) ,where the accuracy of commutation signals significantly decreases during low-speed startup due to the weak back electromotive force,which can lead to commutation failure,this paper proposes a sensorless control method based on a backpropagation neural network optimized by the hippopotamus optimization ( HO) algorithm. The proposed method takes the three-phase voltage and current signals of the motor as inputs to construct a nonlinear mapping model,enabling accurate estimation of the commutation signals. Simulation results show that the proposed HO-BP neural network sensorless control method performs well in the BLDCM system. It can accurately estimate commutation signals,achieve fast and smooth speed control,and has good dynamic response.

Key words: brushless direct current motor,hippopotamus optimization algorithm,BP neural network,sensorless control