Cusum Techniques for Quality Control

Abstract

Summary Existing cusum methods are reviewed, and their relevance to quality control problems is considered. Extensions to the techniques are proposed, and tables, nomograms and charts provided to facilitate their application to practical situations.

An Examination of the Robustness to Non Normality of the EWMA Control Charts for the Dispersion

“…From the preceding discussion, one can see that all these measures are related to the ARL. However, the sole use of the ARL has been criticized (see, e.g., Barnard, 1959;Bissell, 1969;Gan, 1993Gan, , 1994Woodall, 1983). The disadvantage of the ARL is the skewness of the run length distribution in the in and out of control cases as well as in non normality cases.…”

Dispersion chart for some popular distributions under repetitive sampling

“…Larger ARL values are desirable under the in-control process, whereas smaller ARLs under out-of-control process are better so that we are quickly informed that the process has shifted. The use of ARLs for the process evaluation have been criticized by many authors including (Barnard, 1959;Bissell, 1969;F. Gan, 1994;F.…”

Optimal Cusum Control Chart for Censored Reliability Data with Log-logistic Distribution

“…The design of the conventional CUSUM chart involves evaluation of the control-chart performance based on average run length (ARL) [8,9]. The in-control ARL, which refers to the average number of observations from the incontrol process before a false out-of-control alarm is raised, is a measurement of the false-alarm rate.…”