Abstract

This article introduces modified exponentially weighted moving average (modified EWMA) control chart. The modified EWMA control chart is very effective in detecting small and abrupt shifts in monitoring process mean. This chart is based on modified EWMA control chart statistic, which is a correction of EWMA chart statistic and it is free from inertia problem. The advantage of using modified EWMA chart is its good performance for observations that are autocorrelated or independently normal distributed. The performances of Modified EWMA chart are illustrated along with EWMA chart for the two types of analytical process data. The average run length (ARL) properties of modified EWMA scheme are derived using Markov Chain approach.

Key words: Abrupt change, autocorrelated, average run length, exponentially weighted moving average (EWMA), modified.