Application of Maximum Likelihood Estimation Model in X - MR Control Charts

Abstract

Based on the individual and moving range control charts, the maximum likelihood estimation model is proposed to estimate the time of a step change in process variance. It is manifested by Monte Carlo simulation that the estimator coming from this model approaches the actual time very well. We compare the estimator with the first signal time, and come to the conclusion that the estimator detects the process variation more quickly and exactly. So the proposed change point estimator can be easily implemented to improve production efficiency.