Online Short-Circuit Fault Detector of the Induction Motor Winding Based on MRAS Technique
Résumé
This article proposes a new method to diagnose intern-turn short-circuit (ITSC) faults in the stator winding of an induction motor (IM). The ITSC faults change stator winding resistance values, which can be estimated in real time using a model reference adaptive system (MRAS). However, estimation of the stator resistance in the stationary reference frame may not provide accurate information when an incipient ITSC fault occurred in a given phase. As a result, the classical MRAS is limited to detecting symmetric ITSC faults and cannot detect asymmetric ones. To overcome this limitation, a novel approach based on the differences between the measured stator currents in motor phases and the estimated currents obtained from an IM model is proposed. These differences are multiplied by estimated stator currents and used to develop an adaptive mechanism using proportional integral (PI) regulators in each motor phase. To ensure stability, Popov’s criterion inequality is used in the adaptive rule design method. The proposed approach is stable and capable of detecting and estimating different levels of ITSC, including asymmetric faults, in various motor operating modes. The effectiveness of the proposed approach was verified through simulation and experimental results.