Rapin Sunthornwat, Yupaporn Areepong and Saowanit Sukparungsee

DOI: 10.14456/sjst-psu.2018.93

pp. 885 - 895

Abstract

The efficiency of a process, especially a long memory seasonal autoregressive fractional integral moving average
(SARFIMA) process, has commonly been measured through the quality control chart. In this paper, a generalized long memory
SARFIMA process of the exponentially weighted moving average (EWMA) control chart is carried out and shown. Also,
analytical and numerical average run length (ARL) were designed to measure the efficiency of the EWMA control. Existence and
uniqueness by the fixed point theory are proven for the analytical ARL. Error and convergence of numerical integration equations
are also given for the numerical ARL. The findings indicated that the analytical ARL was evaluated more quickly and accurately
than the numerical ARL. As a result, the analytical ARL is an alternative for measuring the efficiency of an EWMA control chart
over a long memory SARFIMA process.

Keywords

• average run length
• EWMA control chart
• SARFIMA process
• composite midpoint rule
• integral equation