ASQ

Exclusive Content & Downloads from ASQ

A Functional Time Warping Approach to Modeling and Monitoring Truncated Degradation Signals

Summary: [This abstract is based on the authors' abstract.] Degradation signals are sensor-based signals that are correlated with degradation processes of engineering components. This article presents a flexible modeling framework for characterizing degradation signals that can only be observed up to a pre-specified failure threshold. The underlying assumption of this framework is that the engineering components degrade according to a similar trend, referred to as the common shape function, but at different degradation rates. Under this assumption, the degradation signals of different components are synchronized using a random time warping process that transforms the common trend function into degradation processes that progress at varying rates. The primary objective of the authors is to obtain real-time predictions for the residual lifetime of components deployed in the field. In the initial step, the historical degradation signals are used to recover the distribution of the degradation processes under the assumptions of the proposed time warping model. Next, the distribution of the degradation process is updated using the signal(s) of partially degraded component (s). The updated model is then used to predict the residual lifetime distributions of these components. The authors test the performance of their methodology using vibration-based degradation signals from a rotating machinery experiment and simulated degradation signals. Additional information and codes are available as supplementary material online.

Anyone with a subscription, including Site and Enterprise members, can access this article.


Other Ways to Access content:

Join ASQ

Join ASQ as a Full member. Enjoy all the ASQ member benefits including access to many online articles.

  • Topics: Data Quality
  • Keywords: Degradation, Data analysis, Multivariate time series, Residual analysis, Lifetime data, Prediction
  • Author: Rensheng R. Zhou, Nicoleta Serban, Nagi Gebraeel & Hans-Georg Müller
  • Journal: Technometrics