Reliability Growth of Continuous Systems.
Abstract
A three-parameter reliability growth model is developed for continuous systems, which model assumes failure rates are improved only after each failure and times between failures are exponential. The basic model is given by lambda sub n = lambda sub u + alpha n (lambda sub o - lambda sub u) where lambda sub n is the failure rate after the nth failure, lambda sub o is the initial failure rate, lambda sub u is the ultimate failure rate, and alpha is a shape parameter. Both maximum likelihood and Bayesian methods of estimating the parameters from test data were investigated. Use of the model for assessment and prediction are addressed and some comparisons are made with the AMSAA model currently in use. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 08, 1982
- Accession Number
- ADA114099
Entities
People
- Darrell D. Penrod
Organizations
- Auburn University