Bearing Health Assessment Based on Different Types of Masses Voting Match and Mahalanobis Distance under Different Operating Conditions

Abstract

In order to solve the problem about low precision of bearing health assessment model caused by the error of model selection under varying operating conditions, a health assessment method based on different types of masses voting match and Mahalanobis distance is first proposed in this paper. Firstly, the method of different types of masses voting match is first used to determine the health assessment model under different operating conditions. This method combines three algorithms with simple structure, good generalization performance and different matching principles. Then Mahalanobis distance is applied to measure the health of the bearing. The experimental results demonstrate its ideal health assessment performance. Simultaneously, the health assessment method proposed in this paper helps managers to take timely maintenance measures and prevent equipment failure as early as possible.