Condition monitoring of wind turbine bearings progressive degradation using principal component analysis

Abstract

Troubles are very uncertain when a vital component of a machine breaks down. Failures of high-speed shaft bearing (HSSB) in a wind turbine are very expensive since it induces the unplanned shutdown of the electrical energy production. In this sense, a vibration-based diagnosis methodology for wind turbine high-speed bearing is proposed using principal component analysis (PCA), applied to extracted features from the vibration signals. Three domains have been investigated in order to extract the features: time domain, frequency domain, and the time-frequency domain. The effectiveness of these features is quantified by two measures, i.e., monotonicity and trendability. The principal component analysis is used to build a health indicator (HI) to describe the health state of the monitored bearing. The potential of this strategy was confirmed utilizing a real run-to-failure vibration history of an HSSB. The exploratory comes about to appear that the proposed approach can effectively identify an early failure.