A multi-dimensional data-driven method for large-size slewing bearings performance degradation assessment

A large-size slewing bearing is usually used in extremely heavy load conditions, and its reliability plays a critical role in machinery performances. Therefore, an ensemble empirical mode decomposition-principle component analysis (EEMD-PCA) based de-noising and performance degradation assessment method was proposed. Firstly, an improved EEMD-PCA based method was conducted on life cycle vibration signals of slewing bearings for de-noising and reconstruction. Afterwards, the reconstructed signals were processed by the PCA, and continues square prediction error (C-SPE) was introduced to represent the performance degradation feature for performance degradation model establishment. The results show that the proposed method is better in unstable signal de-noising than the EEMD-MSPCA, and the established performance degradation model can accurately explain the slewing bearing performance degradation process, which helps enterprises to achieve active maintenances, and provides a potential for further research such as slewing bearing prognostics.