基于RBF神经网络补偿的PMSM反步控制系统

PMSM Reverse Step Control System based on RBF neural Network Compensation

提出了一种基于非线性控制策略的RBF神经网络补偿和反步控制相结合的方法,克服了传统PID控制对非线性控制的缺点．基于神经网络的控制器很好地实现了线性逼近,反步控制保证了系统具有良好的速度跟踪性能．仿真结果表明,该方法比PID控制更有有效性．

This paper presents new, method combining RBF neural network compensation with reverse step control based on non-linear control tactic. This new method overcomes the disadvantages an the control of traditional PID over nan-linear one. Controller based on neural network realizes linear approach suceessfully, and reverse step control assures that system could have good performance of speed track. Simulation results show that this new method is mace effective than PID control.