摘要
讨论了对平稳自相关过程中出现的较小波动进行监控的一种方法。采用自回归移动平均(ARMA)模型对平稳自相关过程进行适当的拟合,通过计算残差的方法消除过程中的自相关要素,并在此基础上提出对于均值和方差出现的较小波动进行监控的指数加权移动平均(EWMA)控制图的构造。通过与其它几种方法的比较来说明该方法在监控平稳自相关过程时有更好的效率。
This paper discusses a method to monitor the small wave shifts in a stationary autocorrelated process. An auto-regressive moving-average model is used to fit the stationary autocorrelated process. Through the residual calculation which used for eliminating the autocorrelation elements; this paper constructs an auto-regressive moving-average model to monitor the mean and variance of the small wave shift. Compared with other methods, the EWMA method can achieve a better efficiency in monitoring a stationary autocorrelated process.
出处
《工业工程》
2006年第3期80-83,共4页
Industrial Engineering Journal
基金
国家杰出青年基金资助项目(70125004)
国家自然科学基金资助项目(70072029
70572050)
关键词
自回归移动平均模型
残差
指数加权移动平均控制图
auto-regressive moving-average model
residuals
exponentially weighted moving average chart