A study on the performance of Lambda Control Chart for autocorrelated processes

Abstract

With the development of modern techniques of online data collection, more and more processes are observed to be autocorrelated. Applying traditional Shewhart control chart or even some advanced control charts like CUSUM and EWMA on autoregressive processes will be misleading. The most popular method dealing with autocorrelation is to firstly fit an ARMA model to the processes, and then apply control charts on the residuals. However, their performance is not satisfactory. In this paper, we study the Lambda Control Chart, and explore the characteristics and performance of this scheme.