Verification of a Monte Carlo Simulation Method to Find Lower Confidence Limits for the Availability and Reliability of Maintained Systems

Abstract

This thesis determined the feasibility and efficiency of a Monte Carlo method of simulating the lower confidence limits for the availabilities and reliabilities of maintained systems. The steady-state availabilities of single-unit systems and the time-constrained availabilities and reliabilities of two-unit parallel systems were simulated. First, a baseline of true exponentially-distributed Mean Time Between Failures (MTBFs) and Mean Time To Repairs (MTTRs) were simulated using the chi-square distribution. Then other MTBFs and MTTRs were simulated to represent sampling of other systems. The availabilities and reliabilities were found using these simulated MTBFs and MTTRs. Next, simulated availabilities and reliabilities were ordered, and lower confidence limits were found. These lower confidence limit point estimates were compared against the systems' exact availabilities and reliabilities. Lastly, the success of this Monte Carlo method is determined by how well the simulated lower confidence limit availability and reliability point estimates cover the exact availabilities and reliabilities.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1985
Accession Number
ADA163832

Entities

People

  • Karen S. Barland

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Advanced Electronics

DTIC Thesaurus Topics

  • Air Force
  • Biodiesels
  • Complex Systems
  • Computational Science
  • Computations
  • Computer Simulations
  • Computers
  • Confidence Limits
  • Differential Equations
  • Engineering
  • Equations
  • Mathematical Models
  • Monte Carlo Method
  • Probability
  • Random Number Generators
  • Reliability
  • Sampling

Fields of Study

  • Engineering

Readers

  • Computational Modeling and Simulation
  • Inertial Navigation Systems.
  • Regression Analysis.