Enhancing the performance of exponential weighted moving average control chart using paired double ranked set sampling

Abstract

The exponentially weighted moving average (EWMA) control chart is good enough to capture a small/moderate shift in the process-mean as it has the ability to provide good detection of a said shift by using current as well as past information about the process. To enhance the performance of EWMA chart, cost-effective and efficient ranked set sampling (RSS) type methods, named paired double-RSS (PDRSS), extreme paired double-RSS (EPDRSS) and quartile paired double-RSS (QPDRSS) are used to propose EWMA-PDRSS, EWMA-EPDRSS and EWMA-QPDRSS charts. By using the simulation study, the performance of developed control charts is evaluated in terms of average run length (ARL) as well as standard deviation of run length (SDRL) measures. The simulation results revealed that the proposed charts surpass the simple random sampling (SRS) based classical EWMA (EWMA-SRS) chart when capturing a small/moderate shift in the process-mean. An illustrative application is also presented to support the proposed study.

KEYWORDS:

• Average run length
• control chart
• exponentially weighted moving average
• in-control
• monitoring
• out-of-control
• paired double ranked set sampling
• process mean
• quartile paired double ranked set sampling
• simulation

Disclosure statement

No potential conflict of interest was reported by the authors.