基于BP神经网络在线辨识的开关磁阻电机神经网络自适应PID控制

PID Controller of Back Propagation Neural Network Control for SRM Based on BP Neural Network on-line Identification

基于开关磁阻电机的高度非线性的电磁特性,固定参数的PID调节器无法得到理想的控制性能指标,该文提出了一种基于BP神经网络在线辨识的SRM神经网络PID自适应控制新方法。实验结果表明,利用BP神经网络来构成开关磁阻电机的神经网络自适应控制器,不但结构简单,而且能适应环境变化,具有较强的鲁棒性。

Based on SRM high nonlinear electromagnetism characteristics, PID controller with fixed parameter can＇t achieve good performance index. This paper presents a novel approach of back propagation neural network PID adaptive control for SRM based on BP neural network on-line identification. Experiments show that the method useed BP neural network to construct the adaptive controller of SRM has the advantages of simple construction, adaptability and robustness.