Predictive adaptive model based control of a five-phase induction motor drive

This paper presents the analysis for a model reference adaptive system (MRAS)-based sensorless vector control of five-phase induction machine. A linear neural network is designed and trained online by means of back propagation network (BPN) algorithm. Moreover, the neural adaptive model is employed in prediction and simulation modes. The ANN-MRAS-based sensorless operation of a threephase induction machine is well established and the same principle is extended in this paper for a five-phase induction machine. In simulation and experiments hysteresis current controller is utilised to adjust current and speed. The results obtained with prediction and simulation mode are compared on the basis of various parameters. Full decoupling of rotor flux control and torque control is realised in both predictive and simulation mode. The results show that predictive method provides better dynamic performance.