Validation of method of moments for uncertainty propagation in reliability estimation

Mitchell Rausch, Haitao Liao

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The method of moments and probabilistic methods such as simulation are usually employed for uncertainty propagation. In particular, the method of moments has been widely used in reliability engineering to calculate the expected value and variance of a reliability estimate given uncertain parameter estimates. However, applying this method in an attempt to obtain the reliability estimate and its variance sometimes can present significant error. Hence, it is important to understand the applicable conditions that the source of uncertainty must uphold to control estimation error. In this paper, two cases are analyzed. The first case involves the Exponential distribution with the failure rate defined according to the Normal distribution. The second case deals with the Weibull distribution with the parameter estimates following the Uniform distribution. Derivations are provided to quantify the error in approximation.

Original languageEnglish (US)
Title of host publicationProceedings - 13th ISSAT International Conference on Reliability and Quality in Design
Pages112-116
Number of pages5
StatePublished - 2007
Externally publishedYes
Event13th ISSAT International Conference on Reliability and Quality in Design - Seattle, WA, United States
Duration: Aug 2 2007Aug 4 2007

Publication series

NameProceedings - 13th ISSAT International Conference on Reliability and Quality in Design

Other

Other13th ISSAT International Conference on Reliability and Quality in Design
Country/TerritoryUnited States
CitySeattle, WA
Period8/2/078/4/07

Keywords

  • Exponential distribution
  • Method of moments
  • Reliability estimation
  • Uncertainty
  • Weibull distribution

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality

Fingerprint

Dive into the research topics of 'Validation of method of moments for uncertainty propagation in reliability estimation'. Together they form a unique fingerprint.

Cite this