A back-stepping neural network control scheme for PM synchronous motors

Abstract

Focusing on the seriously nonlinear problem and unknown or uncertain parameters, a backstepping control method based on neural networks is proposed to realize the multi-object position control of PM synchronous motors. Neural networks in the scheme are used to solve the contradiction between backstepping control and unmatched conditions of systems. A special weight online tuning method is proposed in this paper, and an off-line training phase is not required. The method does not require the system parameters to be exactly known, and the system is robust. The simulation results show that, the proposed method is effective.