基于偏最小二乘的多特性复杂过程监测方法  

Multi-Feature Complex Process Monitoring Method Based on Partial Least Squares

作　　者：孔祥玉 陈雅琳 罗家宇 周红平 叶兴泰 KONG Xiangyu;CHEN Yalin;LUO Jiayu;ZHOU Hongping;YE Xingtai(School of Missile Engineering,Rocket Force University of Engineering,Xi'an 710025,Shaanxi,China)

机构地区：[1]火箭军工程大学导弹工程学院,陕西西安710025

出　　处：《华南理工大学学报（自然科学版）》2022年第6期100-110,共11页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61673387,61833016);陕西省自然科学基金资助项目(2020JM-356)。

摘　　要：作为一种典型的多元统计分析方法,偏最小二乘(PLS)被广泛用于对关键性能指标的异常监测。然而,在复杂的工业过程中,数据可能具有动态、非线性特性或者同时具有动态和非线性特性,基于PLS的线性模型不适用于该过程,且可能导致较高的故障误报率。为此,文中提出了一种基于偏最小二乘的多特征提取算法。首先,基于动态内PLS模型(DiPLS)提取动态特征,得到质量相关动态子空间和动态残差子空间,并针对动态残差子空间进行PLS回归建模,以进一步分离静态信息的质量相关特征;然后,将残差子空间向高维投影,构建非线性PLS模型,以提取变量的非线性特征;最后,在各特征空间中构造统计量,设计了完整的多特性混合系统过程监测策略。田纳西-伊斯曼过程的实例分析结果表明,文中提出的方法有较快的故障检测速度,能达到较好的故障检测效果。As a typical multivariate statistical analysis method,the partial least squares(PLS)has been widely applied in the anomaly monitoring of key performance indicators.However,the data in complex industrial processes may show dynamic or nonlinear characteristics,or both,so the linear model based on PLS is not suitable for this process and may increase the false alarm rate.Therefore,a multi-feature extraction algorithm based on PLS was proposed.Firstly,the dynamic features were extracted based on the dynamic internal model of PLS to obtain the quality-related dynamic subspace and dynamic residual subspace.The PLS regression modeling was carried out for the dynamic residual subspace to further separate the quality related features of static information.Then,the residual subspace was projected to a high dimension to construct a nonlinear PLS model to extract the nonlinear characteristics of variables.Finally,the statistics was constructed in each feature space,and a complete multi-feature hybrid system process monitoring strategy was designed.The example analysis results of Tennessee-Eastman process show that the proposed method has faster fault detection speed and can achieve better fault detection effect.

关 键 词：偏最小二乘 关键性能指标 多特征提取 过程监测 故障检测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]