Spectral and Statistical Based Modeling for Bearing Damage in Induction Motors

Abstract

This paper presents a new modeling approach using non-linear transformation to represent the changes of the statistical parameters which are dependent on the accelerated aging process. This approach enables the extraction of the statistical character of the bearing damage that has been developed during the accelerated aging process. This proposed transformation approach is applied to assess the statistical properties of the aged case data using the initial case data in Induction motor aging process. It can be included that from the results the non-linear transformation plays a very important role to describe the aging process, and the parameters of the function need to be adjusted reflecting the special condition of the process.research, which is based upon bearing damage by the accelerated aging studies, presents both of the spectral and statistical properties of the bearing damage in aging processes. In this sense, it is defined with the non-linear modeling approach to show the statistical parameter changes, which are depend on the aging properties.