中文摘要：

基于Zhou等提出的加权似然比控制图（WEWMA）,给出3种对应的控制图方案,即基于极大似然估计的控制图方案、基于非线性方程估计的控制图方案和基于相关系数矩阵估计的控制图方法,以解决产品缺陷数不可被精确观测过程的在线监控问题。数值模拟显示,基于相关系数矩阵估计构造的控制图方案表现良好,尤其是在核查人员之间的确存在相关性的时候,有更明显的优势。随机生成一个例子,说明了相关系数矩阵控制图的使用方法。

英文摘要：

To monitor latent Poisson count data,we propose three monitoring schemes,using maximum likelihood estimation,non-linear equation estimation and correlation matrix estimationin this paper.These three schemes can monitor process when the number of defects is unknown.Numerical analysis shows that a LWLR_C（latent weighted likelihood ratio test control chart using Chun＇s estimator）scheme based on correlation matrix of multiple reviewers performs better,especially when the reviewers are correlative.We also use a random example to introduce the usage of the LWLR_C chart.