Skip to main navigation Skip to search Skip to main content

Design of Gamma control charts based on the narrowest confidence interval

  • Beihang University
  • City University of Hong Kong

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In statistical process control, Gamma control charts are often designed for time-between-events (TBEs) monitoring. For Gamma distribution, the traditional equal-tail probability limits lead to the problem that some out-of-control events may occur in higher probability than in-control events. To overcome this problem, Gamma control limits with known parameters are proposed based on the narrowest confidence interval. Simulation and case study show that the proposed Gamma control limits outperform the conventional equal-tail probability limits, since the proposed Gamma control charts have shorter width and are more sensitive to detect the increase of parameter θ than the traditional ones.

Original languageEnglish
Title of host publication2016 International Conference on Industrial Engineering and Engineering Management, IEEM 2016
PublisherIEEE Computer Society
Pages219-223
Number of pages5
ISBN (Electronic)9781509036653
DOIs
StatePublished - 27 Dec 2016
Event2016 International Conference on Industrial Engineering and Engineering Management, IEEM 2016 - Bali, Indonesia
Duration: 4 Dec 20167 Dec 2016

Publication series

NameIEEE International Conference on Industrial Engineering and Engineering Management
Volume2016-December
ISSN (Print)2157-3611
ISSN (Electronic)2157-362X

Conference

Conference2016 International Conference on Industrial Engineering and Engineering Management, IEEM 2016
Country/TerritoryIndonesia
CityBali
Period4/12/167/12/16

Keywords

  • equal-tail probability limits
  • Gamma control charts
  • narrowest confidence interval
  • statistical process control
  • time-between-events

Fingerprint

Dive into the research topics of 'Design of Gamma control charts based on the narrowest confidence interval'. Together they form a unique fingerprint.

Cite this