Chapter 4 - Intelligent Data Analytics for 3-Phase Induction Motor Fault Diagnosis Using Gene Expression Programming (GEP)

Abstract

Fault diagnosis and condition assessment (FDCA) of three-phase induction motor (IM) becomes important with increasing age of machine. Proper FDCA plays a pivotal role in enhancing motor’s operational life, efficiency and reducing catastrophic failures. We propose a realistic FDCA method for external fault identification for three phase IMs using Gene Expression Programming (GEP). The proposed approach is validated on publicly available real fault data. The GEP approach uses RMS values of 3-phase voltages and currents as input variables for identifying six types of faults experienced by IM and one normal operating (NF) condition. We compare performance of our GEP approach against ANN and SVM techniques. We also test our GEP approach for analyzing these fault conditions with perturbations in the operating conditions and with varying loads. Simulation results and comparison against ANN and SVM reveal that GEP approach is superior in terms of analytic accuracy and has lower computational requirements.