基于支持向量数据描述的指数加权移动平均控制图

Support Vector Data Description-based Multivariate Exponentially Weighted Moving Average Control Chart

摘要: 基于支持向量数据描述(Support Vector Data Description，SVDD)的D2控制图的重要特点是对过程数据的抽样分布没有特定的要求。由于D2控制图仅使用当前观测点的值来计算统计量，对过程的小偏移并不敏感，在SVDD模型的基础上，提出了基于D2距离的多元加权移动平均(multivariate exponentially weighted moving average，MEWMA)控制图，用S-MEWMA表示。仿真结果表明，无论过程数据服从正态分布还是非正态分布，S-MEWMA控制图均优于 D2控制图。

Abstract: The D2 control chart based on support vector data description (SVDD) has an advantage that it does not require a known sampling distribution for the process data. However, it uses only the current samples, leading to that it is insensitive to small shifts. To solve this problem, based on SVDD method, a multivariate exponentially weighted moving average (MEWMA) control chart (denoted as S-MEWMA) is proposed in this paper. Simulation results show that the S-MEWMA chart outperforms the D2 control chart no matter whether a process follows a normal or non-normal distribution.