基于模型参考自适应的单神经元PID控制器设计

The Design of Single Neuron PID Controller Based on Model Reference Adaptive

将神经网络和模型参考自适应控制系统相结合,研究基于模型参考自适应理论永磁同步电动机(PMSM)BP神经网络速度辨识的单神经元比例积分微分(PID)速度控制器。该系统基于模型参考自适应理论利用BP神经网络辨识器得到的参数信息,使用单神经元PID控制器取代传统PID控制在线调整权值。根据李亚谱诺夫稳定性理论验证了该控制器的稳定性,仿真与实验研究证明该方法改善了PMSM速度控制系统的动态响应与自适应控制能力。

The neural network and model reference adaptive system is combined to study the permanent magnet syn- chronous motor （PMSM） single neuron proportional integral discrete （PID） speed controller with BP neural network speed identification based on model reference adaptive theory neural network.The system based on the model refer- ence adaptive theory is used to obtain the parameter information of the BP neural network.The single neuron PID controller is used to replace the traditional PID control to adjust the weights.According to Lyapunov stability theory, the stability of the controller is verified by simulation and experiments.The method improves the dynamic response and the speed control system of PMSM.