作者: Cynthia A. Lowry , William H. Woodall , Charles W. Champ , Steven E. Rigdon
DOI: 10.2307/1269551
关键词:
摘要: A multivariate extension of the exponentially weighted moving average (EWMA) control chart is presented, and guidelines given for designing this easy-to-implement procedure. comparison shows that run length (ARL) performance similar to cumulative sum (CUSUM) charts in detecting a shift mean vector normal distribution. As with Hotelling's χ2 CUSUM charts, ARL EWMA depends on underlying covariance matrix only through value noncentrality parameter. Worst-case scenarios show should always be used conjunction avoid potential inertia problems. Examples are illustrate use proposed