Sensorless indirect vector control of an induction motor by ANNs observer and EKF

Abstract

This paper describes the application of ANNs observer based on the extended Kalman filter (EKF) for rotor resistance, Mutual inductance and rotor speed estimation of an induction motor, in order to overcome the constraint of adjustment of the covariance matrices and to improve the estimate of these last three parameters. It is an observer for non-linear systems; it is applied to the indirect field oriented control. The magnetic saturation, the operating temperature and difficulties in using sensors for speed measurement are one of the sources of parameters variations. The indirect field oriented control (IFOC) is highly sensitive to these variations and to the noise. To fix this problem, the artificial neural networks (ANNs) observer based on the extended Kalman filter (EKF) is applied to sensorless parameters estimation of an induction motor. The general structure of the EKF is reviewed and the various systems vectors and matrices are defined. The elements of the covariance matrices are properly selected. The EKF associated to the neural network (EKF-ANNs) trained off line algorithm are used to estimate the rotor resistance, the mutual inductance and the rotor speed. By including these parameters as states variables, the EKF equations are established from a discrete two-axis model of the three-phase induction motor. The results obtained with the proposed observer are more efficient than the results obtained with classical EKF.