Login | Register

Fuzzy logic based assignable cause diagnosis using control chart patterns

Title:

Fuzzy logic based assignable cause diagnosis using control chart patterns

Vijayakumar, Sujikumar (2006) Fuzzy logic based assignable cause diagnosis using control chart patterns. Masters thesis, Concordia University.

[thumbnail of MR20765.pdf]
Preview
Text (application/pdf)
MR20765.pdf - Accepted Version
4MB

Abstract

Control charts are widely used in manufacturing and non-manufacturing processes to monitor process average and to reduce variations in quality characteristic. The variations could be either due to common causes, which are inherent nature of the process and unavoidable, or assignable causes which can be diagnosed for rectification. The state of the process, i.e., whether or not the process is statistically in control, is traditionally judged using control limits and unnatural patterns exhibited on control charts. These unnatural patterns, in addition to help in determining the process state, also provide hints on possible assignable cause(s) whenever the process goes out of control. However, there are certain ambiguities associated with this traditional method, such as judging the process state when a point falls exactly on or very near to control limits and vagueness in interpreting unnatural patterns when multiple patterns co-exist on the control chart and relating them to assignable cause. Fuzzy logic has been proved to be an excellent tool for handling such ambiguities and vagueness by quantifying the uncertainty mathematically. A fuzzy inference engine is developed for X歔 chart, based on a chart pattern-cause relationship network. The domain of assignable causes is categorized based on the nature of the shift they can produce, and accordingly related to chart patterns. Each link in the network is represented by a fuzzy inference system which determines the intensity of each cause in the interval [0-1] based on degree of presence of each pattern. All the evidence supporting each cause from the unnatural patterns are aggregated using fuzzy connective operators (max, algebraic sum) and causes are prioritized accordingly so that when process goes out of control, the investigation can be done for the cause having highest priority. The developed fuzzy inference engine is tested with different combinations of unnatural patterns and the results are compared with manual interpretation of control charts and with results from a control chart software tool (MINITAB)

Divisions:Concordia University > Gina Cody School of Engineering and Computer Science > Mechanical and Industrial Engineering
Item Type:Thesis (Masters)
Authors:Vijayakumar, Sujikumar
Pagination:xviii, 138 leaves : ill. ; 29 cm.
Institution:Concordia University
Degree Name:M.A. Sc.
Program:Mechanical and Industrial Engineering
Date:2006
Thesis Supervisor(s):Demirli, Kudret
Identification Number:LE 3 C66M43M 2006 V55
ID Code:9115
Deposited By: Concordia University Library
Deposited On:18 Aug 2011 18:45
Last Modified:13 Jul 2020 20:06
Related URLs:
All items in Spectrum are protected by copyright, with all rights reserved. The use of items is governed by Spectrum's terms of access.

Repository Staff Only: item control page

Downloads per month over past year

Research related to the current document (at the CORE website)
- Research related to the current document (at the CORE website)
Back to top Back to top