An Improved Independent Increment Process Degradation Model with Bilinear Properties

Zhihua Wang, Jiangming Cao, Xiaobing Ma*, Huayong Qiu, Yongbo Zhang, Huimin Fu, Sridhar Krishnaswamy

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

Degradation analysis can be used to assess the reliability of complex systems and highly reliable products because few or even no failures are expected during their life span. To further previous studies on degradation analysis, an independent increment random process method with linear mean and standard deviation functions is presented to model the degradation procedure. It is essentially a Wiener process method with two different transformed time scales. A one-stage maximum likelihood estimation approach is constructed, and the closed form of the product’s median life and the percentile of the failure time distribution (FTD) are also derived. The proposed method is illustrated and verified in a simulation study and practical degradation analysis for IRLED degradation. The Wiener process model with mixed effects is considered as a reference method. Comparisons show that the difference between the two methods regarding the degradation path is not obvious. However, the estimation accuracy of FTD percentile is significantly enhanced by the proposed model, because a more accurate dispersion can be obtained.

Original languageEnglish (US)
Pages (from-to)2927-2936
Number of pages10
JournalArabian Journal for Science and Engineering
Volume42
Issue number7
DOIs
StatePublished - Jul 1 2017

Keywords

  • Failure time distribution
  • Independent increment process
  • Linear mean function
  • Linear standard deviation function
  • One-stage parameter estimation
  • Performance degradation
  • Wiener process model

ASJC Scopus subject areas

  • General

Fingerprint Dive into the research topics of 'An Improved Independent Increment Process Degradation Model with Bilinear Properties'. Together they form a unique fingerprint.

Cite this