Performance Prediction of Rolling Element Bearing with Utilization of Support Vector Regression

Bearings are customary and significant components in any rotating machinery. For an ideal operational rotating system, bearing assumes an indispensable part. Once a catastrophe happens, it will cause gigantic budgetary incidents and even safety hazards. Therefore, it is required to execute a performance assessment and state prediction adequately. Roused from the ongoing works and progress outline, we propose a one-step-ahead prediction strategy in light of the support vector regression analysis over a degradation indicator. From the obtained signal, time-domain feature was computed with the usage of ensemble empirical mode decomposition (EEMD) strategy and further categorized through k-means. The categorization procedure creates the clusters and healthy data cluster has been taken as the reference for degradation indicator calculation. The same has been anticipated utilizing support vector regression (SVR) with one-step-ahead strategy. The outcomes showed the proficiency is at standard and impressively higher than existing strategies.