Publication Date

2-14-2014

Abstract

This work describes a Bayesian model for assessing the reliability of complex systems using component tests, full system tests, and covariate information. Development of the model focused on understanding the relationship between component reliability and system reliability; and defining this relationship using mathematical expressions. The method in this thesis uses pass/fail data coupled with different levels of prior information about system reliability and covariate information to derive posterior distributions that model component and system reliability. This work provides insights on how the number of components, amount of prior information used, inclusion of covariates and spread of failures across components affects point estimates and densities for system reliability. The methodology in this paper is tested using simulated data weapons system test data.

Degree Name

Statistics

Level of Degree

Masters

Department Name

Mathematics & Statistics

First Committee Member (Chair)

Ronald Christensen

Second Committee Member

Shane Reese

Third Committee Member

Michael D. Sonksen

Language

English

Keywords

Bayesian Reliability

Document Type

Thesis

Download

Share

COinS