Evaluation of Combined Reference Frame Transformation for Interturn Fault Detection in Permanent-Magnet Multiphase Machines

This paper focuses on modeling and experimental validation of a diagnostic fault classification procedure for interturn fault detection in permanent-magnet (PM) multiphase machines designed for fault-tolerant electric drives. The diagnostic procedure is based on the symmetrical component theory and relies upon the combined space vector vector D that gathers information from the two original space vectors obtained with different reference frames. The diagnostic index effectiveness and robustness were also investigated against other fault types such as rotor eccentricities and magnet damage to assess its discrimination capability. The proposed procedure was experimentally evaluated for the interturn fault case on a five-phase PM machine. Experiments were carried out at different speed and load levels, with increasing numbers of short-circuited turns. Both simulation and experimental results demonstrated the feasibility of the proposed diagnostic method.