Diagnosing Simultaneous Faults of Bearing and Misalignment in Induction Motor using Combined Method of Bispectrum Analysis of Vibration Signal and KNN Algorithm

The monitoring system for induction motors (IMs) plays an important role in the majority of industrial plants. Bearing faults and shaft misalignment are common mechanical defects in induction motors. The aim of this paper is to detect simultaneously two common faults in induction motor including bearing defect and shaft misalignment. For this purpose, a test setup consisting of an induction motor coupled to a rotor shaft is designed and tested under different loading conditions and at different speeds. The diagnosis parameters of vibration signal are calculated by conventional signal processing methods as well as bispectrum analysis. Feature extraction and KNN classification techniques are applied to the calculated parameters to provide condition monitoring of the induction motor. The results show that the application of bispectrum analysis along with the conventional signal processing methods improves detecting bearing fault in induction motor and shaft misalignment in the case of single fault as well as multiple simultaneous faults.