Reliability prediction based on degradation modeling for systems with multiple degradation measures

Peng Wang, David W. Coit

Research output: Contribution to journalConference articlepeer-review

83 Scopus citations


This paper describes a general modeling and analysis approach for reliability prediction based on degradation modeling, considering multiple degradation measures. Previous research has been focusing on reliability prediction based on degradation modeling at the component failure mechanism level. In reality, a system may consist of multiple components or a component may have multiple degradation paths, so it is necessary to simultaneously consider multiple degradation measures. Also, many research efforts on degradation analysis were initiated by making assumptions about the degradation mechanism. In reality often there is very limited understanding about the degradation mechanism. When the observed degradation data is collected, it is often not known which degradation model is applicable. In this paper, a general analysis procedure is developed. Simulated data were used to illustrate the applicability of this approach. It was verified that, when the multiple degradation measures in a system are correlated, an incorrect independence assumption may underestimate the system reliability.

Original languageEnglish (US)
Pages (from-to)302-307
Number of pages6
JournalProceedings of the Annual Reliability and Maintainability Symposium
StatePublished - May 24 2004
EventAnnual Reliability and Maintainability Symposium - 2004 Proceedings: International Symposium on Product Quality and Integrity - Los Angeles, CA., United States
Duration: Jan 26 2004Jan 29 2004

All Science Journal Classification (ASJC) codes

  • Engineering (miscellaneous)
  • Engineering(all)
  • Safety, Risk, Reliability and Quality
  • Mathematics(all)
  • Computer Science Applications


  • Correlated degradation paths
  • Degradation modeling
  • Multiple degradation measures

Fingerprint Dive into the research topics of 'Reliability prediction based on degradation modeling for systems with multiple degradation measures'. Together they form a unique fingerprint.

Cite this