Integration of Taguchi's loss function with quality control charts

Statistical Process Control has been used broadly in industries, and quality control charts are key tool in SPC, which are using to keep the process under control and continue the process improvement. X-bar is one the common control chart that all the practitioners are applying this chart in to thei...

Full description

Saved in:
Bibliographic Details
Main Author: Rabiee, Ali
Format: Thesis
Published: 2014
Subjects:
Online Access:http://eprints.utm.my/id/eprint/41880/
http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:83051?queryType=vitalDismax&query=Integration+of+taguchi%27s+loss+function+with+quality+control&public=true
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Statistical Process Control has been used broadly in industries, and quality control charts are key tool in SPC, which are using to keep the process under control and continue the process improvement. X-bar is one the common control chart that all the practitioners are applying this chart in to their process improvement. Three parameters of X-bar such as; n, h and k should be determined optimally with regard to the cost parameters of process. The first objective was, to design control chart procedures with an economic approach to consider all cost parameters of process. For covering all the input process cost parameter, Duncan?s cost model was taken as an effective economic model to consider cost parameters and for assessing the loss due to variation in the process; the expected loss in Taguchi's Loss Function for in-control and out-of-control state has been added to the expected period of the in-control and out-of-control state in Duncan's model. The second objective is, to find optimal value of chart parameters. Particle Swarm Optimization (PSO) utilized as a solution algorithm to find the optimal value of economic model. A sensitivity analysis with different value of A, s/?, d, and ? was done to investigate how parameters influence the Expected Total Cost (ETC). Finally, a comparison study with a developed model and reference case reveals ETC in proposed model with the same sample size is smaller than the ETC in reference case