基于神经网络的五相开关磁阻电机直接转矩控制系统研究

以五相开关磁阻电机为研究对象,对直接转矩控制方法进行了分析与研究,提出了一种 BP 神经网络 PID 与传统 PI 复合的新型调速策略.通过高速切换,在大速度误差下采用 BP-PID 控制调速,小速度误差下采用传统 PI 控制调速,使系统误差快速减小并保持稳定,同时在传统 PI 侧引入了以负载转矩差构成的变参数法,可以更好地抵抗外界干扰,提高系统响应速度.对五相开关磁阻电机调速系统进行了仿真实验,实验结果表明,这种控制策略能够有效抑制转矩脉动,调速性能优秀,系统动静态性能良好.

Research on Direct Torque Control System of Five Phase Switched Reluctance Motor Based on Neural Network

Taking five phase switched reluctance motor as the research object , this paper analyzes and studies the direct torque control method , and proposes a new speed control strategy combining BP neural network PID and traditional PI．Through high-speed switching , BP-PID control is adopted for speed control under large speed error , and traditional PI control is adopted for speed control under small speed error ,so that the system error can be quickly reduced and kept stable．At the same time , in the traditional P-PID control , the system error can be reduced and maintained．The variable parameter method based on load torque difference is introduced in I side , which can better resist external interference and improve system responsespeed．The simulation experiment of five phase switched reluctance motor speed control system is carried out．The experimental results show that this control strategy can effectively suppress the torque ripple,it has excellent speed control performance．The system has good dynamic and static performance．