Model-free predictive fault-tolerant control of MPPMSM based on ultra-local model

As the aerospace field continues to develop towards electrification, multi-phase permanent magnet synchronous motors(MPPMSM) have received more and more attention. The fault tolerant control method of MPPMSM based on the traditional model predictive control(MPC) still has the shortcoming of parameter sensitivity, which results in the decrease of the control performance when the parameters of the motor are measured inaccurately or disturbed under different working conditions. In this paper, a model-free predictive fault-tolerant control(MFPFTC) method based on ultra-local model is proposed. By adding an extended sliding mode observer, the prediction errors caused by parameter mismatch are offset. This method improves the parameter robustness of the control method, and also enhances the universality among different parameters motors. The effectiveness of the proposed method is verified by the simulation of open circuit and short circuit fault tolerance based on five-phase H bridge motor.