A dynamic quality control approach by improving dominant factors based on improved principal component analysis

Guangzhou Diao, Liping Zhao and Yiyong Yao

International Journal of Production Research, 2015, vol. 53, issue 14, 4287-4303

Abstract: Process variables in manufacturing process are critical to the final quality of product, especially in continuous process. Their abnormal fluctuations may cause many quality problems and lead to poor product quality. Against this background, this paper proposes a dynamic quality control approach by improving dominant factors (DFs) based on improved principal component analysis (iPCA). Firstly, the generation of iPCA is illustrated to identify the DFs which lead to quality problems. Then, a quality prediction model for improving DFs is proposed based on modified support vector machine (SVM). An incremental weight is introduced in SVM to improve its sparsity and increase the accuracy of quality prediction. Thus, the product quality can be guaranteed by controlling the DFs dynamically. Finally, a case study is provided to verify the feasibility and applicability of proposed method. The research is expected to provide some guidance for continuous process.

Date: 2015
References: Add references at CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://hdl.handle.net/10.1080/00207543.2014.997400 (text/html)
Access to full text is restricted to subscribers.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:taf:tprsxx:v:53:y:2015:i:14:p:4287-4303

Ordering information: This journal article can be ordered from
http://www.tandfonline.com/pricing/journal/TPRS20

DOI: 10.1080/00207543.2014.997400

Access Statistics for this article

International Journal of Production Research is currently edited by Professor A. Dolgui

More articles in International Journal of Production Research from Taylor & Francis Journals
Bibliographic data for series maintained by Chris Longhurst ().