Development of a methodology for bearing fault scrutiny and diagnosis using SVM

摘要

In recent days, the necessity to condition monitoring and maintenance has increased to a greater extent. Even though there are various main components available in the induction motor, bearing is something crucial where we need to concentrate. Hence the fault occurs in the bearing has to be paid more intention. In most of the cases, holes had been formed in the bearing of the induction motor for fault analysis. Whereas considering the industrial environment, the chance of occurrence of scratch is considerably higher than the possibilities of holes. Hence in this paper, scratch has been considered as a factor for fault analysis. Thus by considering the two types of scratches on the outer raceway of bearing, experimental results were obtained at the different load condition. Following that, the detection of the bearing fault is achieved by the Fast Fourier Transform. Finally, the diagnosis was performed with the help of Support Vector Machine.