Abstract
The ability to accurately track the degradation trajectories of rotating machinery components is arguably one of the challenging problems in prognostics and health management (PHM). In this paper, an intelligent prediction approach based on asymmetric penalty sparse decomposition (APSD) algorithm combined with wavelet neural network (WNN) and autoregressive moving average-recursive least squares algorithm (ARMA-RLS) is proposed for degradation prognostics of rotating machinery, taking the accelerated life test of rolling bearings as an example. Specifically, the health indicators time series (e.g., peak-to-peak value and Kurtosis) is firstly decomposed into low frequency component (LFC) and high frequency component (HFC) using the APSD algorithm; meanwhile, the resulting non-convex regularization problem can be efficiently solved using the majorization-minimization (MM) method. In particular, the HFC part corresponds to the stable change around the zero line of health indicators which most extensively occurs; in contrast, the LFC part is essentially related to the evolutionary trend of health indicators. Furthermore, the nonparametric-based method, i.e., WNN, and parametric-based method, i.e., ARMA-RLS, are respectively introduced to predict the LFC and HFC that focus on abrupt degradation regions (e.g., last 100 points). Lastly, the final predicted data could be correspondingly obtained by integrating the predicted LFC and predicted HFC. The proposed methodology is tested using degradation health indicator time series from four rolling bearings. The proposed approach performed favorably when compared to some state-of-the-art benchmarks such as WNN and largest Lyapunov (LLyap) methods.
Highlights
Rotating machines are critical elements of almost all forms of mechanical assemblies, which play an important role in today’s industrial applications
Most developments in the mechanical fault prognostics area have been targeted towards directly utilizing degradation-based data to trace the degradation trajectories, very few studies have used the idea of sparse decomposition
The original health indicators od degradation trajectory are rearranged as two components, i.e., low frequency component (LFC) and high frequency component (HFC)
Summary
Rotating machines are critical elements of almost all forms of mechanical assemblies, which play an important role in today’s industrial applications. Both the convex and nonconvex methods shrink all the coefficients and remove too much energy of useful information, resulting in the separation of the LFC and HFC becoming more challenging To overcome these limitations and to robustly separate the LFC and HFC from degradation trajectories of health indicators, a novel asymmetric penalty sparse decomposition (APSD) algorithm with non-convex sparsity constraint is proposed in this work. In an effort to establish a comprehensive assessment of degradation processes, the proposed fusion prognostics framework combines the WNN (a nonparametric-based approach) and the ARMA-RLS (a parametric-based approach) in an attempt to improve the prediction performance, in which the physical information of the degenerative process and massive sample data are not considered and accommodated. To address the issue of models merging and improve the prediction accuracy, the health indicators time series (HITS) is decomposed into different sub-components, i.e., low frequency component (LFC) and high frequency component (HFC), using the APSD algorithm.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
