A General Approach for Current-Based Condition Monitoring of Induction Motors

Abstract

The development and validation of a novel current-based induction motor (IM) condition monitoring (CM) system is described. The system utilizes only current and voltage signals and conducts fault detection using a combination of model-based and model-free (motor current signature analysis) fault detection methods. The residuals (or fault indicator values) generated by these methods are analyzed by a fuzzy logic diagnosis algorithm that provides a diagnosis with regard to the health of the induction motor. Specifically, this includes an indication of the health of the major induction motor subsystems, namely the stator windings, the rotor cage, the rolling element bearings, and the air-gap (eccentricity). The paper presents the overall system concept, the induction motor models, development of parameter estimation techniques, fault detection methods, and the fuzzy logic diagnosis algorithm and includes results from 110 different test cases involving four 7.5 kW four pole squirrel cage motors. The results show good performance for the four chosen faults and demonstrate the potential of the system to be used as an industrial condition monitoring tool.