Change point detection in process control with robust individuals control chart

Abstract

A crucial part of the inference of industrial process change point detection is that once a suspected change point is detected, the statistical test needs to be carried out to ensure the significance of the change. Conventional test statistics such as Maximum-Type testing method is very sensitive to outliers as the formulation of the estimator is lack of robustness. In addition, oftentimes, evidence of larger shifts could be mistakenly downweighted or rejected stemming from these outliers and also due to the designated score function. The proposed robust individuals control chart adopting the Huber Maximum-Type (HMT) testing method is based on a well-motivated Huber score function, which appears to provide protection against this complication. Extensive numerical simulation studies were carried out to assess the performance of the proposed procedure with its counterparts. Taking a real data set from industry, we illustrate the usefulness and applicability of the proposed chart in practice.