In this paper, two attribute control charts under multiple dependent state sampling (MDS) and its modification are proposed to monitor the multivariate Poisson count data. The sensitivity of the MDS is increased by introducing an additional parameter. The new method is named as generalized multiple dependent state sampling (GMDS). The proposed charts are based on two pairs of limits i.e. inner and outer control limits that make use of the past samples information in addition to the current sample. The monitoring statistics of the proposed charts are the sum of non-conformities or flaws of each quality characteristics that make the charting structure relatively simple. The performance comparison of the proposed charts with the existing multivariate Poisson control chart is made by using out-of-control average run length (ARL). The comparison reveals that the ARLs of the proposed charts are considerably smaller than the existing scheme for all shift sizes and different values of parameters. Simulation study also showed the dominance of the proposed charts over other existing control charts in detecting the process shift. For practical application, an empirical illustration is provided to signify the performance evaluation of the proposed charts.

本文提出了两个在多重相关状态采样（MDS）下的属性控制图及其修改，以监视多元Poisson计数数据。通过引入其他参数，可以提高MDS的灵敏度。该新方法被称为广义多相关状态采样（GMDS）。建议的图表基于两对限制，即内部和外部控制限制，除了当前样本外，这些限制还使用过去的样本信息。提议的图表的监视统计信息是使图表结构相对简单的每个质量特征的不合格或缺陷之和。通过使用失控平均游程长度（ARL）可以将建议的图表与现有的多元Poisson控制图进行性能比较。比较表明，对于所有班次大小和不同参数值，建议图表的ARL都比现有方案小得多。仿真研究还表明，在检测过程偏移时，建议的图表比其他现有的控制图表更占优势。对于实际应用，提供了经验说明以表示对建议图表的性能评估。