基于PSO-ENN动态神经网络的无刷直流电机控制

Brushless DC Motor Control Based on PSO-ENN Dynamic Neural Network

无刷直流电机转子位置检测属于电机控制中的一项关键技术,针对无位置传感器的无刷直流电机,其线反电动势过零点即为电机的实际换向点,传统采用的BP神经网络算法建立线反电动势与换向信号之间的模型存在响应速度慢、开关误导通现象。因此,采用Elman动态神经网络建立两者之间的函数模型,可提高控制系统的动态特性,并采用粒子群算法优化Elman网络的初始权值与阈值,进一步提高网络的收敛速度与计算精度。仿真结果表明,与传统的BP网络控制算法相比,新算法能更好地实现电机的精确换向,系统动态性能更好、控制精度更高、鲁棒性更强。

The rotor position detection of brushless DC motor is a key technology in motor control, for the sensorless BLDC motor, the zero crossing point of the line back EMF is the actual commutation point of the motor.The traditional BP neural network algorithm has slow response speed and a misleading switch between the line back EMF and the commutation signal. Therefore, Elman dynamic neural network is used to establish a functional model between them, which can improve the dynamic characteristics of the control system, and use particle swarm optimization to optimize the initial weights and thresholds of Elman network, further improving the convergence speed and accuracy of the network. The simulation results show that compared with the traditional BP network control algorithm, the new algorithm can better achieve precise commutation of the motor, and the system has better dynamic performance, higher control accuracy and stronger robustness.