Lower Confidence Limit for Reliability Based on Grouped Data with a Quantile Filling Algorithm

Abstract

The purpose of this article is to derive a lower confidence limit for reliability given a grouped data set. This is done by using a quantile filling algorithm which generates pseudo failure data from grouped data. A general framework of this approach is first introduced. The cases for the exponential distribution, Weibull distribution, and lognormal distribution are used to illustrate this approach. Simulation studies are carried out and the results show that it is a useful method handling grouped data. Two field lifetime data sets are also analyzed to demonstrate its feasibility. Some further improvements are discussed as well.