| Summary: | To handle the limitations from two types of methods, a hybrid prognostics approach is put forward to predict the remaining useful life (RUL) of wind turbine bearings in this paper. Firstly, the features of the signal are extracted from the time domain and the frequency domain. Secondly, features with better monotonicity are selected through the Spearman rank correlation analysis method and the hierarchical clustering method. Thirdly, a feature fusion method is applied to fuse the selected features into health indicator (HI) based on the percentage of the total variance. Finally, the RUL is achieved by the exponential degradation model. Moreover, the position of the first prediction is detected by applying the T-test. Implementation results indicate that the proposed approach is practical and effective in the RUL prognostics of wind turbine bearings.
|
|---|
