非正态总体下的EWMA控制图  

EWMA Chart for Non-normal Distribution

作　　者：吴卫东[1] 王海宇[2] 

机构地区：[1]延安大学管理学院,延安716000 [2]中原工学院经管学院,郑州450007

出　　处：《世界科技研究与发展》2011年第4期700-702,748,共4页World Sci-Tech R&D

基　　金：国家杰出青年基金项目(70125004);国家自然科学基金项目(70572050)

摘　　要：主要讨论了非正态分布情况下EW。MA图的效率分析和优化设计问题。首先引入赋权方差法将非正态总体转换为正态过程,由此构造适用于非正态总体的EWMA图,然后以平均产品长度作为效率度量对非正态EWMA控制图进行了优化设计。通过与其他几种方法的比较说明该方法在监控非正态总体中出现的较小波动时具有的优势。The problem of efficiency analysis and optimal design for a EWMA chart in non- normal distribution are discussed. First, by introducing Weighted Variance method,a non-normal distribution is transformed to normal distribution. From this method, the EWMA chart for non-normal distribution is applied and constructed. Then, the EWMA chart for non-normal distribution is designed according to efficient measurement of average run length. By comparing with other control charts, it is clearly revealed that this method has great advantages in detecting little drift for non-normal distribution.

关 键 词：EWMA 赋权方差法 平均运行长度 

分 类 号：O213.1[理学—概率论与数理统计]