基于概率SVDD的控制图研究  

Research on Control Chart Based on Probability SVDD

作　　者：程旭超 Cheng Xu-chao(School of Science,Yanshan University,Qinhuangdao 066004,Hebei Province,China)

机构地区：[1]燕山大学/理学院,河北秦皇岛066004

出　　处：《科学与信息化》2023年第12期91-93,共3页Technology and Information

摘　　要：基于支持向量数据描述算法,构建一种有效的指数加权移动平均控制图,用于监控高维数据的生产情况。通过模拟仿真,对所提出的控制图及原有控制图的性能进行研究分析。模拟结果表明,基于概率支持向量数据描述算法的指数加权移动平均控制图在检测小偏移方面比原有控制图更有效。Based on the support vector data description algorithm,an effective exponential weighted moving average control chart is constructed to monitor the production of high-dimensional data.Through simulation,the performance of the proposed control chart and the original control chart is studied and analyzed.The simulation results show that the exponential weighted moving average control chart based on probabilistic support vector data description algorithm is more effective than the original control chart in detecting small deviation.

关 键 词：支持向量数据描述 指数加权移动平均控制图 平均链长 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]