Studies in Reliability Economics and Reliability Modeling and Inference
Abstract
Research accomplishments under ARO support can be grouped into seven problem areas. In *reliability economics*, we solved the important open problem of identifying optimal system designs relative to criterion functions based on both performance and cost. We also solved the related statistical estimation problem, allowing data-based approximations of optimal system designs. New results were obtained in *structural reliability*, and a research monograph in this area was completed. Sufficient conditions were obtained for the *optimal allocation* of the reliabilities p(1)....p(n) to the n components of a coherent system. Nonparametric inference under type I censoring has been developed for the distribution of bug discovery times in *software reliability*. Work on *systems with weighted components* clarifies their place within the theory of coherent systems and sheds light on their monotonicity properties. We introduced the concept of "self-consistency" in *Bayesian inference* and demonstrated its relevance to the "consensus problem". New inference procedures for dealing with autopsy data from *stress-strength experiments* have been justified.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 25, 2008
- Accession Number
- ADA500005
Entities
People
- Francisco J. Samaniego
Organizations
- University of California