Diagnosis of switch open-circuit fault in PM brushless DC motor drives

An adaptive neuro-fuzzy inference system (ANFIS) is developed to diagnose open switch faults in the inverter bridge of PM brushless DC motor drives. Performance of the drive system under normal and faulty conditions is obtained through a discrete-time model. The motor DC-link current is monitored over one electrical cycle under healthy and faulty operations. The time-domain waveform is processed using wavelet transform, and suitable indices are derived to train ANFIS. Testing of the diagnosing system shows the effectiveness of the proposed methodology.