A comparative study on non-linear state estimators applied to sensorless AC drives: MRAS and Kalman filter

In this paper, two different nonlinear estimators applied to sensorless AC drives, Kalman filtering techniques (EKF and UKF) and model reference adaptive system (back emf and reactive power models), are discussed and compared to each other. Both of the observer types are studied and analyzed both experimentally and theoretically. In order to compare the observers precisely, the observers are tested under the identical conditions.