基于核典型相关性-熵成分分析的工业过程质量监测方法  

A quality monitoring method for industrial process based on kernel canonical correlation-entropy component analysis

作　　者：彭开香[1] 张丽敏[2] PENG Kai-xiang;ZHANG Li-min(School of Automation,University of Science and Technology Beijing,Beijing 100083,China;Key Laboratory of Knowledge Automation for Industrial Processes of Ministry of Education,Beijing 100083,China)

机构地区：[1]北京科技大学自动化学院,北京100083 [2]工业过程知识自动化教育部重点实验室,北京100083

出　　处：《控制与决策》2021年第12期2999-3006,共8页Control and Decision

基　　金：国家自然科学基金项目(61873024)。

摘　　要：工业过程多变量、数据高维度和非线性的特点使得对其质量监测及质量相关的故障诊断变得复杂.融合核熵成分分析(KECA)及典型相关分析(CCA)方法的思想,进行特征提取降维的同时确保所提取特征与质量变量的最大相关性,提出一种新的质量相关的工业过程故障检测方法.首先,采用KECA对输入数据进行核空间的映射及特征提取,同时融合CCA算法思想使得所提取特征与质量变量间关联最大化;然后,构建监测统计量并用Parzen窗估计其控制限,用于过程的故障检测;最后,运用所提方法对带钢热连轧工业过程实际生产数据进行分析,并与其他4种传统非线性算法对比分析,实验结果验证了所提方法的准确性、有效性及先进性.The characteristics of industrial process such as multivariate,high-dimensional data and nonlinearity complicate the quality monitoring and quality-related fault diagnosis.In this paper,we present a novel quality-related fault detection method for industrial process by combining the kernel entropy composition analysis(KECA)and the canonical correlation analysis(CCA)algorithms for feature extraction,which reduces the number of input space dimension and ensures the maximum correlation between the extracted features and quality variables simultaneously.Firstly,the KECA algorithm is used to extract the features of the standardized data,the idea of the CCA algorithm is used to maximize the correlation between the extracted features and quality variables.Then,the monitoring statistics are constructed for process failure detection and the control limits are estimated via invoking a Parzen window density estimator.The proposed method is applied to the actual data of hot strip mill process(HSMP).Comparing with the performance of other four classical algorithms which are also suitable for nonlinear data,the results verify the accuracy,efficiency and advance of the method proposed.

关 键 词：质量监测 故障诊断 带钢热连轧 核熵成分分析 典型相关分析 PARZEN窗 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]