基于神经网络的开关磁阻电机自适应控制

针对开关磁阻电机显著的非线性特性，将具有非线性映射能力及自适应能力的误差反向传播（BP）神经网络应用于开关磁阻电机驱动系统（SRD），并结合传统比例、积分和微分（PID）控制的优点，提出一种基于BP神经网络的开关磁阻电机在线辨识与自适应PID控制方法．该方法利用BP神经网络实时观测系统输出，优化PID控制参数，对于解决开关磁阻电机由于非线性严重而导致控制困难的问题具有较强的针对性．实验结果证明了该方法的有效性，且系统适应性强，稳定性好，响应速度和控制精度均令人满意。

Adaptive Control for Switched Reluctance Motors Based on Neural Network

Aiming at the nonlinear electromagnetic characteristic of switched reluctance motor(SRM),the strong points of back propagation (BP) neural network and traditional proportional,integral and differential (PID) controller are combined,and then a new control solution,adaptive PID control and on-line identification based on BP neural network,is presented.This method uses BP neural network,which has powerful nonlinear mapping and self-adaptation capabilities,to realize real-time observation of the output of the system,and on- line adjustment to the parameters of PID controller.With the proposed method,satisfactory response speed and precision as well as good robustness and stable performance were obtained by experiments.