Fuzzy Diagnostics of Rotor Bar Breakage in Induction Motor Based on Time Synchronous Averaging and Wavelet Transform

This study addresses the detection of broken rotor bar (BRB) faults in induction motors under diverse load conditions, including no-load scenarios, which pose significant challenges due to spectral leakage effects. The proposed method integrates time synchronous averaging (TSA) and discrete wavelet transform (DWT) to extract and analyze fault characteristics from the residual stator current. The sixth-level DWT approximation coefficient a6 is employed as an energy-based fault severity index, effectively isolating fault-related harmonics. Experimental validation on a 1.1 kW squirrel cage induction motor demonstrates the method’s efficacy, with slip variations recorded between 0.0232 and 0.0963 under varying load conditions. Results indicate a clear distinction in a6 energy values, particularly under no-load (0%) and full-load (100%) conditions, with significant improvements in fault detection accuracy after applying TSA. A fuzzy inference system further enhances the classification of fault severities, achieving reliable identification of motor states ranging from healthy to severe faults. Compared to existing techniques, the proposed approach demonstrates superior performance in low-load conditions and reduces computational complexity, making it suitable for real-time diagnostics. This work establishes a robust framework for BRB fault diagnosis, contributing to the reliability and predictive maintenance of induction motors.