Jiang, Wei; Tsui, Kwok-Leung; Woodall, William H. (2000, ASQ and American Statistical Association) Dept. of Industrial Engineering & Engineering Management, The Hong Kong Univ. of Science & Technology; School of Industrial & Systems Engineering, Georgia Institute of Technology; Dept. of Statistics, Virginia Inst. of Technology & State University
[This abstract is based on the author's abstract.] A new control chart is proposed. The autoregressive moving average (ARMA) chart is based on monitoring an ARMA statistic of the original observations. The special cause chart (SCC) of Alwan and Roberts and the EWMAST chart of Zhang are special cases of the ARMA chart. Simulation studies show that the ARMA chart is competitive to the optimal exponentially weighted moving average chart for iid observations. It is better than the SCC and EWMAST charts for autocorrelated observations. An informal procedure for determining the appropriate parameter values of the proposed chart is developed. This procedure is based on two signal-to-noise ratios. Two real examples are given to demonstrate advantages of the new chart.
Process analysis,Quality control (QC),Simulations,Statistics,Average run length (ARL),Control charts