A bivariate exponentially weighted moving average control chart based on exceedance statistics. (January 2023)

• Record Type: Journal Article
• Authors: Mahmood, Tahir; Erem, Aysegul
• Is Part Of: Computers & industrial engineering. Volume 175(2023)
• Publication Date: 2023-01
• Subjects: Control chart -- Location monitoring -- Order statistics -- Real-time monitoring -- Statistical process control; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108910
• Languages: English
• ISSNs: 0360-8352

Abstract

Abstract: Nonparametric control charts are more practical tools for statistical process control (SPC), as they are robust in situations in which the underlying distribution is unknown. Comprehensibility and simplicity of exceedance statistics provide great convenience to analysts in multivariate SPC applications. By using the exceedance statistics, analysts save time and avoid complex calculations. Therefore, in this study, a bivariate nonparametric exponentially weighted moving average (BEWMA-EX) control chart is proposed based on the exceedance statistics to detect the shifts in the location parameter. The performance of the proposed BEWMA-EX chart is compared with the multivariate sign EWMA (MSEWMA) control chart under some well-known bivariate distributions, such as bivariate normal, t, and gamma distributions. The BEWMA-EX chart outperforms the MSEWMA control chart in terms of run-length properties. To highlight the importance of the stated study, the BEWMA-EX chart is implemented on industrial engineering datasets related to coal power plant and aluminum electrolytic capacitor manufacturing processes. The findings are promising and support the simulated results. Highlights: Proposing a distribution-free BEWMA control chart. The BEWMA chart relies on the exceedance statistic. The performance of BEWMA-EX chart is provided under some well-known bivariate distributions. A comparative study with the existing MEWMA chart is provided. Implementation of BEWMA-EX chart on industrial engineering datasets.