联合检测过程位置和尺度参数的非参数控制图

发布时间：2019-05-13 16:18

【摘要】：传统质量控制过程中,经常分别对属性特征的均值和方差进行监测,并且大多数控制图都需要假设受控分布是正态的.但实际中,由于没有足够多的数据来估计过程的分布.因此,受控分布很少能精确知道.而非参数控制图对过程分布不需做任何假设.本文我们提出一个新的基于Cram′er-von Mises(CvM)检验的非参数指数加权移动平均控制图,称为ECvM图,用于联合监测过程位置参数和尺度参数.结果显示,与已有的非参数控制图相比,多数情况下本图具有良好的性能,文中通过模拟给出了对于给定的平均运行长度ECvM图的控制限,并通过对一些位置参数和尺度参数的模拟得到了该图的步长分位数、平均运行长度以及步长标准差来说明该图的受控和失控性能,结果显示,ECvM图对于各种漂移情况都有很好的性能,最后用一个实例来说明ECvM图的实际应用.
[Abstract]:In the process of traditional quality control, the mean and variance of attribute features are often monitored respectively, and most control charts need to assume that the controlled distribution is normal. In practice, however, there is not enough data to estimate the distribution of the process. Therefore, the controlled distribution is rarely accurately known. The nonparametric control chart does not need to make any assumptions about the process distribution. In this paper, we propose a new nonparametric exponential weighted moving average control chart based on Cram'er-von Mises (CvM) test, which is called ECvM chart, which is used to jointly monitor the position parameters and scale parameters of the process. The results show that compared with the existing nonparametric control diagrams, this graph has good performance in most cases. In this paper, the control limit of the given average running length ECvM diagram is given by simulation. Through the simulation of some position parameters and scale parameters, the step size quantile, average running length and step size standard deviation of the graph are obtained to illustrate the controlled and runaway performance of the graph. ECvM diagram has good performance for all kinds of drift cases. Finally, an example is given to illustrate the practical application of ECvM graph.
【学位授予单位】：辽宁大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O213.1

【相似文献】