This thesis is concerned with the development of a computer-aided Life Data Analysis System (LIDAS) on personal computers. LIDAS analyzes the life data of products or parts and gives estimates of various characteristics of the life distribution such as its parameters, reliability function, hazard rate function, MTTF/MTBF, etc. The types of life data LIDAS can analyze are complete data, singly censored data, multiply censored data, repairable and unrepairable series system test data, and after-service data.
Graphical and maximum likelihood methods are used to estimate the parameters of life distribution in parametric analysis. The distributions considered are exponential, two and three parameter Weibulls, extreme value, normal, and longnormal distributions. LIDAS gives some informations to help a user to choose an appropriate distribution; sum of squared error in graphical estimation and Kolmogorov-Smirnov type maximum deviation statistic in maximum likelihood estimation. In non-parametric analysis, the Kaplan-Meier and Herd-Johnson methods are used to estimate the reliability functions.
LIDAS is interactive menu-driven system written in BASIC programming language and runs on all IBM compatible PC/XT/AT personal computers.