Optimal design of AEWMA control chart with new sampling strategy

The application of engineering control strategy and high sampling rates lead to the increase of autocorrelation of process data, which will inflate the average run length (ARL) of control chart and deteriorate its sensitivity to the occurrence of assignable cause. In this paper, we investigate the optimal design of AEWMA control chart with mixed sampling strategy and s-skip sampling strategy. These sampling strategies will reduce the autocorrelation within a sample. The adaptive algorithm will overcome the inertia problem of EWMA control chart. A Markov chain method is given to calculate the ARL of AEWMA control chart. Numerical analysis shows that the mixed sampling AEWMA chart outperforms the s-skip sampling AEWMA chart for high levels of autocorrelation. The analysis result shows that AEWMA control chart outperforms Shewhart X control chart implementing these sampling strategies.