Adaptive Neuro-Fuzzy Control for Electric Machine Emulation of PMSM Under Fault

Accurate fault modeling and real-time simulation of faulty electric machines are essential for developing fault-tolerant control strategies in power electronics and drive systems. This paper presents a simulation-based evaluation of an adaptive neuro-fuzzy inference system (ANFIS)-based control approach for permanent magnet synchronous motor (PMSM) fault emulation under inter-turn short-circuit faults (ISTCF) and stator resistance deviation (SRD) conditions. The proposed method uses an adaptive neuro-fuzzy controller, enabling dynamic compensation for fault-induced disturbances. Conventional PI-based controllers struggle with dynamic fault transitions and system uncertainties, resulting in poor tracking accuracy and slower fault adaptation. The ANFIS controller, in contrast, adjusts its membership functions and inference rules in real time, improving fault compensation, current tracking accuracy, and stability. Simulation results demonstrate that the ANFIS-based approach significantly reduces steady-state error and the divergence index (DI) compared to PI control, ensuring better adaptability to fault conditions. The results highlight ANFIS as a promising alternative for real-time fault-tolerant motor control, with future work focusing on expanding the fault models and integrating the proposed approach into real-time hardware testing environments.