生活用纸质量监控与异常分析方法的研究

Study on Quality Monitoring and Abnormal Analysis Method for Tissue Paper

关键词:  生活用纸  多变量统计过程控制  质量分析  异常识别

Key Words:tissue paper  multivariate statistical process control  quality analysis  abnormal identification

摘要:某生活用纸制造企业以单变量标杆值来检测产品质量合格与否，未考虑各变量之间的相关性，该种检测方法必将导致误判率提高。如今国内外已将统计过程控制模型应用在质量监控中，但是生活用纸包含多种质量变量，仅对某一单独变量进行监控显然不够。基于此本课题提出多变量统计过程控制模型，该模型在单变量统计过程控制（SPC）基础上从整体角度对产品质量进行监控。并将该模型运用在江门某生活用纸制造企业的产品质量监控与分析上，找出质量异常的产品，并结合综合变量贡献率来分析异常原因。最后通过引入10组故障点验证模型的准确度，结果表明，本课题研究方法有效可行。

Abstract:The quality of tissue paper directly affect its sales market. A tissue paper mill tests the value of a single variable to determine whether qualified the product is without taking into account the integrity of the variables, this test method will lead to increase of misjudgment rate. Nowadays, the statistical process control model has been applied in quality monitoringat at home and abroad, tissue paper contains a variety of quality variables, evidently only a single variable monitoring is not enough. Therefore, the Multivariable Statistical Process Control （MSPC） model was proposed in this paper, which monitored the product quality from the overall perspective on the basis of SPC. And the model was applied to the product quality monitoring and analysis in a tissue paper mill in Jiangmen, Guangdong Province, to find out the abnormal product and analyze the abnormal reason with the contribution rate of the integrated variable. Finally, the accuracy of the model was verified by introducing 10 sets of fault points. The results showed that the MSPC method was effective and feasible for quality monitoring．