Statistical process control of stationary processes

Abstract

The burnt-in rate for simulated stationary and invertible ARMA1, 1(u, v) time series is discovered to be [u^(2t)]*(v a?? 1)^2. Thereafter, the first comprehensive simulation study is done to evaluate the performance-effort tradeoff between optimal EWMAST and optimal ARMAST control charts. The optimal EWMAST control chart is adequate when u is not high positive with small mean shift.