A Combined Bayes-Sampling Theory Method for Monitoring a Bernoulli Process
Abstract
The author assumes a population of one-shot missiles which are stored in a ready or near ready state at the physical point of their deployment. He hopes that the missiles will sit in idle waiting for many years, but this allows environmental effects to degrade the missile's capability of successful deployment. Since even a brand new missile may fail to operate properly, and there are no important physical differences between the individual missiles in the given population, he assumes that a randomly selected missile will have a probability Pt of successful deployment, or reliability, at time t. It is obviously important to monitor Pt, so a sample of the missiles is tested periodically. Since the testing is destructive, the population is eventually depleted by the testing. Furthermore, defects in missile design may be uncovered, so modifications may be introduced which will have a tendency to increase the reliability. For technical reasons, however, the author chose to describe a test which is designed to detect a deterioration in the reliability.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1984
- Accession Number
- ADP003840
Entities
People
- N. D. Singpurwalla
- R. L. Launer
Organizations
- United States Army Research Laboratory