Evaluation of Rolling Bearing Performance Degradation Based on Comprehensive Index Reduction and SVDD

The evaluation of rolling bearing performance degradation has important implications for the prediction and health management (PHM) rotating equipment. A method based on comprehensive index reduction support vector data description (SVDD) is proposed in this study. Firstly, improved variational mode decomposition (VMD) was used to decompose vibration signals, defect frequency amplitude ratio which sensitive early faults extracted. Secondly, a feature set bearings constructed by combining traditional time-domain time–frequency-domain indexes, main features are extracted dimensionality algorithm locally linear embedding (LLE). Finally, SVDD model utilized characterize evaluate lifetime process using distance from test sample trained hypersphere center. Results showed that can accurately detect occurrence weak fault stage objectively reveal bearings.