Bearing Failure Prognostic Method Based on High Frequency Inductance Variation in Electric Railway Traction Motors

Abstract

This paper deals with the analysis of a prognostic technique able to analyze the health condition of traction motor bearings. While the methods commonly adopted in literature use vibration and acceleration signals, the proposed method is sensor-based and entirely based on an electromagnetic approach. The bearing conditions are monitored through the variation of the value of a high frequency inductance coil positioned near the bearing. The wear, the corrosion and the defects influence the magnetic behavior of metal parts of the bearing and therefore the total value of the coil inductance. The measurement data are processed in a regression model. The experimental results show the feasibility of the proposed prognostic technique.