永磁交流伺服系统的神经网络自适应控制研究

Research on the Intelligent Adaptive Control Based PMSM Servo System

针对永磁交流伺服系统,提出了一种自适应 PID 控制器学习算法——带预测模型的神经网络PID 控制方法。该方法采用一个三层 RBF 网络辨识交流伺服系统的特性,用另一个 BP 神经网络作为自适应控制器。仿真结果表明,在系统参数发生变化和存在负载转矩扰动的情况下,该方法具有较好的自适应能力和良好的动态性能。

In view of PMSM servo system, a learing algorithm of self-adaptive PID control neural networks with model of prediction is presented in this paper. A three layers of RBF neural network is used with another BP is as self-adaptive controller to obtain the model of PMSM servo systems Simulation results show that the proposed controller has good self-adaptability and excellent dynamic performance while the parameter variation and load torque disturbance occurs.