Demerit-GWMA Control Chart for Demerit Statistics

Abstract

A classical Demerit control chart is used to monitor the process through which various types of defects in complex products, such as automobiles, computers, mobile phones, etc. are found in general. As a technique for rapidly detecting small shifts of the process mean in the control chart, the EWMA(exponentially weighted moving average) technique is very effective. This study suggested the Demerit-GWMA control chart, combining the GWMA(generally weighted moving average) technique, which shows better performance than EWMA technique in detecting small shifts of process mean, into the classical Demerit control chart, and evaluated its performance. Through the evaluation of its performance, it was found that the Demerit-GWMA control chart is more sensitive than both the classical Demerit control chart and the Demerit-EWMA control chart in detecting small shifts of process mean.

Generally weighted moving average monitoring schemes: Overview and perspectives

“…A demerit scheme is used to monitor a demerit system containing different types of defects in an inspection unit. For instance, defects are classified in terms of categories or classes, for example, “Very serious,” “Serious,” “Moderately serious,” and “Minor.” Kang et al 39 proposed the GWMA demerit scheme to monitor the mean demerit number per unit by classes. It is shown that, in many situations, the GWMA demerit scheme has a better OOC performance than the Shewhart and EWMA demerit schemes.…”

“…Kang and Baik 73 proposed the DGWMA demerit scheme and they showed that, in many situations, it has better performance in detecting small shifts in the mean demerit number per unit by classes than the GWMA, EWMA, and Shewhart schemes discussed in Kang et al 39 …”

Demerit-GWMA Control Chart with Fast Initial Response

“…Dodge found that defects of the complex industrial products such as automobiles did not occur only in a type, and in order to address this problem, he suggested classifying them into four classes as follows depending on the extent to which they affect various types of defects. Because such a Demerit control chart uses 3σ Center line and Control limits, it not only requires the normality assumption on the statistics, but it also is not sensitive to small shifts of the process mean [2,3,7].…”

“…The Demerit-GWMA control chart was first suggested by H.W. Kang et al [7] in 2011. Demerit-GWMA statistics( i W ) are defined as in the following equation ( 6), in a similar manner to GWMA statistics ( i G ).…”

“…Recently, H.W. Kang et al suggested the Demerit-GWMA control chart, which shows better performance in detecting small shifts in process mean by extending the EWMA technique [7].This study suggested the FIR Demerit-GWMA control chart, in which the FIR feature is integrated to the Demerit-GWMA control chart and the performance of those control charts was comparatively evaluated.…”