Fault-Tolerant Control Using the GA Optimization Considering the Reluctance Torque of a Five-Phase Flux Switching Machine

This paper deals with the fault tolerance of a five-phase flux switching machine. Short-circuit currents calculation considering inductances variation is developed. Machine behavior (torque quality, copper losses, and homopolar current) under a single short-circuit phase fault, two consecutive and nonconsecutive phases short-circuited, is simulated with a two-dimensional finite elements (2-D FE) model and validated experimentally. Then, a new method is developed to improve its performances in faulty mode, by reconfiguring reference currents. In fact, an accurate torque model is established and then used in a genetic algorithm to optimize reference currents in faulty mode. In this approach of reference currents computation, the used algorithm has multiobjectives and multiconstraints, thereby allowing choosing the appropriate fault-tolerant current solution according to our application. The torque model is considered to be more accurate and closer to the 2-D FE results in both healthy and faulty modes. Then, a comparison of machine performances in healthy, degraded, and reconfigured modes is presented. Experimental results corroborate the analysis.