Spectral and Discriminant Analysis Based Classification of Faults in Induction Machines

Abstract

This paper presents a new condition indicator for classifying of stator and rotor related faults in induction motors. It relies on the characteristic fault frequencies of the motor in question and can be extended to different types of motors with different magnetic structures. The proposed method, occupied band-power ratio, focuses on the power concentration of the characteristics fault frequencies and yields the final result as a unit-less quantity. Features developed using this method are studied using linear data explanatory tools and further optimized with Discriminant Analysis for classification. The efficacy of the proposed method is validated experimentally by using grid and inverter fed induction motors.