Reliability Estimation in Series Systems

Download

Download PDF View on GitHub

Abstract

This paper develops maximum likelihood estimation techniques for reliability analysis of series systems with Weibull-distributed component lifetimes under right-censoring and masked failure data. A likelihood ratio test is developed for selecting between full heterogeneous-shape and reduced homogeneous-shape models, demonstrating proper Type I error control and good power to detect meaningful departures from shape homogeneity.

Keywords

Series Systems Weibull Distribution Maximum Likelihood Masked Data Right-Censoring Bootstrap Likelihood Ratio Test Model Selection