ON CONFIDENCE LIMITS FOR THE PERFORMANCE OF A SYSTEM WHEN FEW FAILURES ARE ENCOUNTERED

Abstract

In some situations encountered today the components are so reliable that no failures are observed within the time available for testing. This can pose a problem in data analysis and interpretation. Consideration is given to the problem of determining lower confidence bounds on the reliability of a complex system, such as the Saturn 1-C, when each component is assumed to have an exponential life and different components have different multiplicities within the system. We discuss and compare the confidence limits obtainable from various interpretations of the data and several theories such as asymptotic likelihood ratio, asymptotic maximum likelihood, the Bayesian method and the new method presented here utilizing the conditional probabilities of malfunction of the components given that a malfunction in the system has occurred.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1967
Accession Number
AD0668983

Entities

People

  • Sam C. Saunders

Organizations

  • Boeing

Tags

Communities of Interest

  • Air Platforms
  • Weapons Technologies

DTIC Thesaurus Topics

  • Complex Systems
  • Computational Science
  • Computations
  • Confidence Limits
  • Data Analysis
  • Data Science
  • Engineering
  • Failure Mode And Effect Analysis
  • Information Science
  • Probability
  • Random Variables
  • Reliability
  • Statistical Analysis
  • Statistical Inference
  • Statistics
  • Stochastic Processes
  • Time Intervals

Fields of Study

  • Mathematics

Readers

  • Aerospace Test and Evaluation
  • Regression Analysis.
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms