A Comparative Study on Adaptive EKF Observers for State and Parameter Estimation of Induction Motor

In this article, conventional extended Kalman filter (EKF) and adaptive extended Kalman filters (AEKFs) based on adaptive fading, strong tracking, and innovation are compared for state and parameter estimation problem of induction motor (IM) by considering their estimation performances and computational burdens. The estimation performance of EKFs depends on the proper selection of system and measurement noise covariance matrices. However, it is hard to select optimum elements of those matrices using the trial-and-error method, and those are affected by the operating conditions of IM. Therefore, different AEKF approaches with the ability to update those matrices online according to the operating conditions have been proposed in the literature. However, to the best of the author's knowledge, no comparison has been yet reported as to which observer is more effective for real-time state and parameter estimation problem of IM. This paper focuses on the detailed comparison of those observers and provides useful results to the literature.