ANN-based system for inter-turn stator winding fault tolerant DTC for induction motor drives

Direct Torque Control (DTC) scheme uses the stator resistance of the machine for the estimation of stator flux. The variation of stator resistance, due to stator winding turn fault, creates an error in the estimated stator flux position that may subsequently cause a failure of the complete drive system. This paper proposes the possibility of developing a remedial operating strategy using artificial neural network (ANN), which ensures a fault tolerance of inter-turn stator winding fault in the direct torque control (DTC) for induction motor drives. The proposed fault tolerant approach is achieved using a strategy that detects inter-turn stator winding fault, identifies fault severity and improves the DTC performance in the presence of incipient stator winding turn fault. The fault tolerant system is obtained by tuning the stator resistance to make the DTC strategy more robust and precise. This allows continuous disturbance-free operation of the induction motor drives even with existing inter-turn stator winding faults. This strategy is simple to implement, does not require new sensors or changes in the standard drive system. Experimental implementation is demonstrated for the validity of the proposed idea.