A control scheme for monitoring process covariance matrices with more variables than observations. (1st October 2018)

• Record Type: Journal Article
• Authors: Li, Zhonghua; Tsung, Fugee
• Is Part Of: Quality and reliability engineering international. Volume 35:Number 1(2019)
• Page Start: 351
• Page End: 367
• Publication Date: 2018-10-01
• Subjects: high dimension -- individual observation -- process variability -- quality control -- statistical process control; Reliability (Engineering) -- Periodicals; Quality control -- Periodicals; High technology -- Periodicals; 620.00452
• DOI: 10.1002/qre.2403
• Languages: English
• ISSNs: 0748-8017

Abstract

Abstract: In this paper, we propose a new control chart that integrates a powerful high‐dimensional covariance matrix test with the exponentially weighted moving average procedure for monitoring high‐dimensional variability with individual observations. Design and implementation of the proposed chart are provided, including search algorithm and a table for the control limits, diagnostic aids after the signal, effect of misspecifying the in‐control distribution, and a bootstrap procedure. Monte Carlo simulation results show that the new chart, with its powerful inherited properties, provides satisfactory performance in various cases, especially for covariance shifts that involve diagonal components. The application of the proposed method is illustrated with a real data example from a white wine production process.