DEVELOPING MIXED EXPONENTIALLY WEIGHTED MOVING AVERAGE-CUMULATIVE SUM & CUMULATIVE SUM EXPONENTIALLY WEIGHTED MOVING AVERAGE CONTROL CHARTS

Control charts are powerful tools of statistical process control (SPC) used for monitoring processes. Control charts are not only used to detect changes in process, but also enable the quality engineers to prepare the system for an appropriate response before production of the faulty products. The main objective of most researches on control charts is developing a favorable and cost-eective design as well as most appropriate sampling methods. In most of the researches that have been done in this eld, the control charts are designed based on simple random sampling (SRS). However, there are many charts in the literature of statistical process controlthatarenotbasedonSRSandusealternativemethods such as ranked set sampling (RSS) or generalized methods. The RSS method is an alternative to SRS when taking actual measurement of the quality characteristic is costly or production of the faulty products is expensive. In this paper, in order to improve the performance of mixed EWMA-CUSUM (MEC) and mixed CUSUM-EWMA(MCE)controlchartsinestimatingthe population mean, ranked set sampling method rather than common simple random sampling method is used to design these mixed control charts. The performances of the proposed MEC-RSS and MCE-RSS control charts are evaluated through simulation studies using the average run length (ARL) criterion. Moreover, the performances of the proposed control charts are compared withthoseofCUSUM-SRS,CUSUM-RSS,EWMA-SRS, EWMA-RSS, MEC-SRS, and MCE-SRS control charts. The results revealed that RSS method improves control charts performance compared to SRS method because it providesanecientestimationofthemeanandvariance ofthepopulation. Inaddition, usingtherankedsetsampling scheme makes the proposed mixed control charts moresensitivetodetectingthesmallandmoderateshifts in the process mean than the other corresponding classical control charts. Moreover, based on the same sampling methods, the performances of mixed control charts are better than those of the single CUSUM and EWMA control charts.