Effective utilization of signals collected by distributed sensor networks is crucial for tracking degradation and forecasting the remaining useful life (RUL) of rolling bearings. The phase space warping (PSW) algorithm constructs the hierarchical dynamics to physically describe damage evolution. However, the PSW algorithm is unable to handle multivariate signals. To enable synchronous tracking of degradation in multivariate signals, the proposed solution is the multivariate phase space warping (MPSW) algorithm. First, the multivariate signals are embedded in the reconstructed phase space. Second, the local polynomial receives the current phase space trajectory (PST) to predict the reference PST, after which damage indicators are extracted by comparing the current PST with the reference PST. Third, robust principal component analysis with tensor smooth constraint was proposed on the DIs tensor to extract the main degradation pattern. Finally, the degradation is input to the exponential degradation model to predict the RUL. The run-to-failure experimental datasets for rolling bearings are applied to validate the effectiveness of the proposed MPSW. Experimental results demonstrate that the proposed MPSW effectively tracks the multivariate degradation, and accurately predicts the RUL with distributed sensor networks.
13 October 2022 in IEEE Transactions on Instrumentation and Measurement
Predictive maintenance plays a crucial role in the field of intelligent machinery fault diagnosis, which improves the efficiency of maintenance. This paper focuses on the extraction of real-time damage feature and the prediction of remaining useful life (RUL) in predictive maintenance of rolling bearings. Some RUL prediction approaches lack dynamic foundations and require large amounts of data and prior knowledge. This paper proposes the algorithm of segmented relative phase space warping (SRPSW) and a strategy combining double exponential model (DEM) and particle filter (PF) to predict the RUL. SRPSW provides a dynamic basis for real-time RUL prediction in different stages. The DEM-based PF reduces the need for prior knowledge and improves accuracy. The analysis results from normal and accelerated degradation experiments show that the proposed SRPSW overcomes the failure of the original PSW in depicting the later operating stage of bearings. Further, the relative damage indicators (RDIs) extracted by SRPSW are more accurately than commonly used indicators. The predicted results show that the DEM-based PF does not require professional and prior information whilst ensuring the accuracy of RUL prediction. The proposed approach in this paper provides a new avenue for predictive maintenance of bearings.
Fuzzy dispersion entropy (FuzzyDE) effectively solves the dispersion pattern rounding errors of dispersion entropy (DE). However, distinct pattern characteristics correspond to different properties of dynamic systems. The pattern probability derived from merely counting the number of different patterns imposes limitations on the representation of nonlinear systems by FuzzyDE. To address this issue, this paper presents feature fuzzy dispersion entropy (FFuzzyDE). This approach utilizes the numerical characteristics of dispersion patterns as new indicators for measuring pattern probability, thereby replacing the singular quantification methods of traditional approaches. The fuzzy membership is combined with the characteristics of the patterns to derive the exclusive probability associated with each dispersion pattern. Simulation experiments demonstrate that the feature dispersion pattern construction scheme allows FFuzzyDE to achieve stable entropy values more rapidly across varying data lengths, is sensitive to characteristic changes in frequency-amplitude modulation signals, effectively captures subtle characteristic changes in nonlinear systems, and exhibits excellent noise suppression capabilities. Furthermore, by integrating the amplitude-based time-shift coarse-graining method, FFuzzyDE is extended into the multiscale calculation domain, resulting in amplitude-based time-shift multiscale feature fuzzy dispersion entropy (ATSMFFDE). Two sets of bearing fault damage experiments show that ATSMFFDE can accurately extract bearing fault characteristics, and therefore provides effective health monitoring indicators for Aero-engine bearings.