Performance analysis and comparison of two fault-tolerant model predictive control methods for five-phase PMSM drives

Abstract

Model predictive current control (MPCC) and model predictive torque control (MPTC) are two derivatives of model predictive control. These two control methods have demonstrated their strengths in the fault-tolerant control of multiphase motor drives. To explore the inherent link, the pros and cons of two strategies, the performance analysis and comparative investigation of MPCC and MPTC are conducted through a five-phase permanent magnet synchronous motor with open-phase fault. In MPCC, the currents of fundamental and harmonic subspaces are simultaneously employed and constrained for a combined regulation of the open-circuit fault drive. In MPTC, apart from the torque and the stator flux related to fundamental subspace, the x-y currents are also considered and predicted to achieve the control of harmonic subspace. The principles of two methods are demonstrated in detail and the link is explored in terms of the cost function. Besides, the performance by two methods is experimentally assessed in terms of steady-state, transition, and dynamic tests. Finally, the advantages and disadvantages of each method are concluded.