# Robustness of reliability predictions based on the exponential component failure distribution

- UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
- Susan Payne Varner (Creator)
- Institution
- The University of North Carolina at Greensboro (UNCG )
- Web Site: http://library.uncg.edu/
- Advisor
- William Powers

**Abstract:** The reliability of a component at a specified time, t, denoted by R(t), is defined as the probability that the component survives at least until time t. The reliability depends on the failure distribution which governs the component failure and, in most situations, the exact form of the component failure distribution is unknown. Traditionally, the exponential component failure distribution has been the most widely used component failure distribution. This is largely due to two reasons. First, the calculations involving the exponential component failure distribution are relatively simple to perform, thereby making it desirable to use from a mathematical viewpoint. Second, a large body of theoretical results have been derived about the exponential component failure distribution, thus reinforcing its popularity. The use of the exponential component failure distribution as the assumed component failure distribution results in the equivalent assumption that the failure rate of the component is constant. That is, given that a component has survived for t0 units of time, the conditional probability that the component survives for a given period of time A is independent of t0. Obviously, in the real world, most components do not have a constant failure rate.

Robustness of reliability predictions based on the exponential component failure distribution

PDF (Portable Document Format)

6709 KB

Created on 1/1/1975

Views: 27

## Additional Information

- Publication
- Thesis
- Language: English
- Date: 1975
- Subjects
- Exponential families (Statistics)
- Distribution (Probability theory)
- Principal components analysis

Email this document to