基于MPCA建模的间歇过程故障分析  

Fault Analysis of Batch Process based on MPCA Modeling

作　　者：张静静 方彬 秦兰 刘佳微 张博健 ZHANG Jing-jing;FANG Bin;QIN Lan;LIU Jia-wei;ZHANG Bo-jian(Hebei Institute of Architecture and Civil Engineering,Zhangjiakou,Hebei 075000)

机构地区：[1]河北建筑工程学院,河北张家口075000

出　　处：《河北建筑工程学院学报》2022年第1期169-173,共5页Journal of Hebei Institute of Architecture and Civil Engineering

基　　金：河北省教育厅自然科学基金青年项目(QN2022108)。

摘　　要：检测具有复杂性的间歇过程是非常困难的,它具有复杂性。为了最终达到对间歇过程的更有效地检测和诊断,降低系统的误报率,本文以pensim青霉素发酵过程的仿真为基础,设计了一种以多向主元分析为基础的故障检测方法,并利用MPCA和聚类分析法对间歇时段的故障进行检测。It is very difficult to detect the batch process with complexity.In order to detect and diagnose the batch process more effectively and reduce the false alarm rate of the system,based on the simulation of Pensim penicillin fermentation process,a fault detection method through multidirectional principal component analysis is designed,and MPCA and cluster analysis are used to detect the faults in the batch period.

关 键 词：间歇过程 聚类分析 故障检测 主元分析 

分 类 号：TP15[自动化与计算机技术—控制理论与控制工程]