Using Multivariate Tukey distribution in Multivariare Process Capability Indices

Please enable javascript in your browser.

Ebrahimi, Samaneh | 2013

621 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 44434 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Mahlooji, Hashem
Abstract:
In this thesis we will provide a comprehensive study of process capability indices, and we use multivariate g-and-h Tukey distribution for calculating process capability indices of multivariate non-normal processes. Univariate process capability indices have been used commonly in literature for measuring the capability of a process. However, there are several processes with more than one quality characteristic. When these characteristics are independent, we can use univariate process capability indices, but when the characteristics are dependent we should use multivariate methods for measuring process capability. There have been several studies about multivariate process capability indices, but most of these studies are based on the assumption of normality of the process. However, in real world there are plenty of processes when the underlying distribution is non-normal. For non-normal multivariate processes there have been few studies regarding process capability. Therefore, in this study we will concentrate on using the multivariate Tukey g-and-h distribution in order to measure the process capability of multivariate non-normal quality characteristics
Keywords:
Process Capability ; Multivariate Tukey Distribution ; Multivariate Process ; Non-Normal Distribution

No TOC