Implementation ofNeuralNetworksBased Speed ControllerinDirect Torque Control ofInduction Motor Drive
This paper represents performance evaluation of two different speed controller for Voltage Source Inverter(VSI) fed three phase Induction Motor(IM) drives which operates based on Direct Torque Control(DTC) Topology. The proposed Neural Network(NN) speed controller which operates online monitoring based on Adaptive Linear Network(ADALINE) to improve the performance of IM drive system. NN based speed controller is trained in closed loop system with various data under dynamic conditions. The proposed NN based model has better accuracy in speed and torque tracking capability with respect to reference values compare to conventional PI controllers. The proposed speed controller is needed to produce reference torque for DTC. In order to update the weights in ADALINE model the speed values are continuously monitored and the values are automatically updated with respect to time. By using MATLAB simulation results neural based speed controller is more effective as compared to classical PI based speed controller during load changing conditions.