基于神经网络变结构控制的PMSM系统研究

The Study of PMSM System Base on Neural Network & Variable Structure Control

针对交流永磁同步电机伺服系统中,变结构控制策略存在的抖动,采用递归神经网络来调节变结构控制器参数,有效的削弱了变结构控制固有的抖动特性。仿真结果表明,该方案不仅具有良好的动态性能,而且对系统参数变化和负载扰动具有很强的鲁棒性。

Aimed at the dithering of variable structure control（VSC） AC PMSM servo system, weakened the intrinsic dithering of VSC by adopting recursive neural network to adjust it＇s parameters. The simulative result illuminate that the control strategic has good dynamic performance and great robustness against parameter variations and load disturbances.