基于改进的PSO-BP神经网络的无刷直流电机控制

实时检测无位置传感器的无刷直流电机运行过程中的转子位置,并输出相应的开关管导通与关断信号是控制无刷直流电机的一项关键技术.本文针对神经网络控制存在开关管误导通的问题,引入非线性的惯性权重因子并采用异步时变的学习因子改进策略对标准粒子群算法(PSO)进行改进,通过改进的PSO算法对BP神经网络进行优化,进而控制无位置传感器无刷直流电机的换相顺序.仿真实验表明,采用基于改进的PSO-BP神经网络方法控制无刷直流的运行,可以取得满意的结果.

Control of brushless DC motor based on improved PSO BP neural network

Real-time detection for rotor position during the operation of sensorless brushless DC motor and outputting a corresponding switch signal is a key technology for controlling the brushless DC motor.In order to overcome the problem of neural network exists the local optimal value,the nonlinear inertia weight factor is introduced and the learning factor is improved by adopting the improved strategy of asynchronous time variation to improve the standard particle swarm optimization (PSO).Then the improved particle swarm optimization algorithm combined with BP neural network is used to control the commutation order of sensorless brushless DC motor.Simulation results show that the control of brushless DC motor based on improved PSO-BP neural network can achieve satisfactory results.