统计与决策2025，Vol.41Issue(8)：47-52,6.
基于贝叶斯检验动态预测置信区间的统计过程控制方法
A Statistical Process Control Method Based on Bayesian Test for Dynamic Prediction Confidence Interval
摘要
Abstract
Time series data processing is an important direction of multivariate statistical process control method research.PCI is a method for detecting abnormalities in time series data by predicting confidence intervals.This paper uses the Bayes test to correct the prediction results and do adaptive processing,proposing a statistical process control model based on Bayes-PCI,and then through numerical simulation,makes a comparison on the average running length under different confidence p and window size W,verifying the performance of the proposed method in processing one-dimensional and high-dimensional data,respective-ly.Finally,the paper employs the process data of Tennessee-Eastman to prove the feasibility of the proposed method in practical application.关键词
统计过程控制/Bayes检验/置信区间/平均运行长度Key words
statistical process control/Bayes test/confidence interval/average running length分类
数理科学引用本文复制引用
朱永忠,夏世源..基于贝叶斯检验动态预测置信区间的统计过程控制方法[J].统计与决策,2025,41(8):47-52,6.基金项目
国家自然科学基金重点项目(71831006) （71831006）
