A robust sliding-mode control strategy of Permanent Magnet Synchronous Motor

This paper applies a global sliding mode variable structure control method based on neural network reaching law to Permanent Magnet Synchronous Motor. Parameters, ε and K, which were determined previously in the conventional reaching law, are regulated adaptively by two back-propogation neural networks (bpNN) respectively. Thus, a new global sliding mode control scheme is presented. Simulation results show that a desired performance is achieved; meanwhile system state is always sliding on the sliding surface, so system chattering is eliminated and good robustness is obtained.