CUSUM控制图的一些改进及应用

发布时间：2018-03-04 08:08

本文选题：统计过程控制　切入点：CUSUM控制图　出处：《四川师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：控制图是对生产过程进行质量控制的一类重要工具,CUSUM控制图是控制图理论中非常重要的一类控制图,它是由序贯概率比检验推得的,具有较好的统计性质.针对用CUSUM控制图进行过程监控时遇到的一些问题,本文做了以下三方面工作:第一,CUSUM控制图对现实生产过程进行监控时,在参数选取上需要根据实际情况进行斟酌才能达到较好的监控效果.文中结合电动汽车电机性能的检测数据,说明如何应用CUSUM控制图对生产过程进行在线监控.有效地利用CUSUM控制图对电动汽车电机性能检测数据进行在线监控,能够规避传统检测方法的不足,及时给出生产过程发生系统误差的预警信息,提前做好检修,降低产品损失率.第二,应用CUSUM控制图解决实际问题时,需要对监控的数据提前做出准确的分析,拟合数据潜在的分布形式,预估可能发生漂移的大小,而这些分析往往比较困难.文中引入自适应非参数CUSUM (ANC)控制图,使得在对过程进行监控时省去拟合参数分布与估计漂移大小的步骤,实现对生产过程的有效监控.同时以对电气产品绝缘性能数据的监控过程为例,阐述ANC控制图使用方法及监控效果.第三,为了使控制图在对过程进行监控时,既能够监控各种分布形式的数据,又能够监控不同大小的漂移,同时还能根据当前样本信息动态决策下一步如何抽样,提高监控效率.文中结合变采样区间技术构造VSI-ANC控制图,详细说明该控制图参数的设置与选择,给出一些参数设置参照表;并将VSI-ANC控制图与ANC控制图进行比较,说明监控效果;同时将VSI-ANC控制图应用到经典实例中进行监控原理的说明与监控效果的验证.
[Abstract]:Control chart is an important tool for quality control of production process. CUSUM control chart is a very important control chart in control chart theory. It is derived from sequential probability ratio test. Aiming at some problems encountered in the process monitoring with CUSUM control chart, this paper has done the following three aspects of work: first, when the control chart of CUSUM is used to monitor the actual production process, In order to achieve better monitoring effect, the selection of parameters needs to be considered according to the actual situation. In this paper, the performance of electric vehicle motor detection data, This paper explains how to use CUSUM control chart to monitor the production process on line, and how to use CUSUM control chart to monitor the performance data of electric vehicle motor effectively, which can avoid the deficiency of traditional detection method. The early warning information of the system error in the production process is given in time, the overhaul is done well in advance, and the loss rate of the product is reduced. Secondly, when using the CUSUM control chart to solve the practical problems, we need to make an accurate analysis of the monitored data in advance. It is difficult to fit the potential distribution of data and estimate the magnitude of possible drift. In this paper, adaptive nonparametric CUSUM ANC control chart is introduced. When monitoring the process, the steps of fitting the parameter distribution and estimating the drift are omitted to realize the effective monitoring of the production process. At the same time, the process of monitoring the insulation performance data of electrical products is taken as an example. The use method and monitoring effect of ANC control chart are expounded. Thirdly, in order to make the control chart monitor the process, it can not only monitor all kinds of distributed data, but also monitor the drift of different sizes. At the same time, it can also dynamically decide how to sampling according to the current sample information to improve the efficiency of monitoring. This paper constructs VSI-ANC control chart combined with variable sampling interval technology, and explains in detail the setting and selection of the parameters of the control chart. Some parameter setting reference tables are given, and the VSI-ANC control chart is compared with the ANC control chart to illustrate the monitoring effect. At the same time, the VSI-ANC control chart is applied to the classical examples to explain the monitoring principle and verify the monitoring effect.
【学位授予单位】：四川师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O213.1

【参考文献】