Speed Control Of DFIM Using Artificial Neural Network Controller

Abstract

Nonlinear characteristics and parameters variation of the Doubly Fed Induction Motor (DFIM) posed a serious problem during operation. For this purpose, it is necessary to use control laws insensitive to variations in parameters, disturbances, and non-linarites. In this paper, a speed controller of a DFIM by the application of a PI controller based on Artificial Neural Network (ANN) is proposed. The results obtained with ANNPI are compared with AFLC-PI. This controller is then designed and trained online using a back propagation network algorithm. The performance of the proposed controller is adopted using Matlab / Simulink. Simulation results show a fast dynamic response and good performance in tracking speed and torque.