Discrete Reliability Growth Models Using Failure Discounting.

Abstract

Three discrete reliability growth models using fractional failure reduction, referred to as failure discounting, were developed to estimate changing system reliability. Each of the models is designed for use when testing is performed until a fixed number of failures have been observed and attribute data, success or failure, is available for each trial. The first reliability growth model applies failure discounting to the maximum likelihood estimate for a proportion. The second and third models use a modification of an exponential reliability estimate employing linear regression and a weighted average technique respectively along with failure discounting to track changing reliability. Two failure discounting methods were used with each reliability growth model. The first method reduces past failures by a fixed fraction at a fixed interval. The second method uses the upper confidence bound for the reoccurrence of each failure cause as the discounted failure value. The performance of the reliability growth models with varying reliability growth patterns was evaluated with a Monte-Carlo computerized simulation. (Theses)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1987
Accession Number
ADA186380

Entities

People

  • James E. Drake

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Computations
  • Computer Programs
  • Computer Simulations
  • Computers
  • Confidence Limits
  • Data Science
  • Engineering
  • Estimators
  • Failure Mode And Effect Analysis
  • High Reliability
  • Monte Carlo Method
  • Probability
  • Random Variables
  • Reliability
  • Simulations
  • Statistics

Fields of Study

  • Engineering

Readers

  • Life Cycle Cost Analysis
  • Statistical inference.