Reliability Estimation in the Presence of Censored Data A Case Study in an Automotive Part Manufacturing System

Please enable javascript in your browser.

Seyed Esfahani, Zahra | 2013

745 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 43960 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Eshraghniaye Jahromi, Abdolhamid
Abstract:
In this study a brief definition of the basic concepts of the reliability and censoring data is given as well as the literature review of the censored data and introduction to concept of robust. In order to find suitable distribution by considering censored data, the goodness of fit test, Chi-squared test, and Anderson-Darling approaches are used. Moreover to estimate Parameter of the reliability distribution, Maximum Likelihood estimation (MLE) is given. A case study in an automotive part manufacturing system is presented and the distributions of reliability parameters of several critical components (i.e. reliability, hazard rate, etc.) have been estimated. Considering all data, and omitting outliers, illustrated that the robust procedure extracted more precise results
Keywords:
Invariance ; Failure Analysis ; Censored Data ; Robust Estimator

No TOC