Performance Degradation Assessment for Bearing Based on Angle-Time Cyclostationarity

Rolling element bearing signals are known to exhibit pseudocyclostationary (pseudo-CS) properties that limit the efficiency of health monitoring. This study extends angle-time CS (AT-CS) to performance degradation assessment (PDA) of rolling bearings. First, the life-cycle phase map is proposed to demonstrate the degradation process and the cyclic kinematic properties of energy distribution. Second, combining the entropy feature extraction technique of nonlinear dynamics, the AT instantaneous power spectrum entropy (AT-IPS-E) and AT IPS permutation entropy (AT-IPS-PE) are further introduced to extract energy distribution characteristics. Finally, the feature vector set (FVS) composed of AT-IPS-E and AT-IPS-PE is constructed as the bearing health indicators (HIs), and comparative experiments are conducted by assessment models. The analysis results from the two bearing fault degradation examples demonstrate the validity and reliability of the FVS as a new HIs, while also confirming the universality of the PDA model.