Change Points Estimation in Shewhart Control Charts Using Fuzzy Clustering Approach

Control charts are mainly used to determine whether a process is in a state of statistical control or not. If the chart shows the out-of-control signal, it indicates a variation in the process. Aweakness of these charts is that they can not indicate the real time of the process change to determine the source of variations.In this article we investigate the problem of finding change points in different types of control charts by modifying a fuzzystatistical clustering algorithm. For implementing the algorithm in change points estimation, we considered the probability of membership of each observation to its assumed cluster and use itas a similarity measure in clustering. Using an objective function, process change points in different types of control charts wereestimated with either fixed or variable sampling strategy. Several simulation runs are conducted to evaluate the performance of the proposed approach for some types of control charts.

Keywords:

Statistical Process Control , Control Chart , Fuzzy Clustering , Change Points