Mixture Bayesian Regularization of PCR Model and Soft Sensing Application

被引:57
作者
Ge, Zhiqiang [1 ]
机构
[1] Zhejiang Univ, Dept Control Sci & Engn, State Key Lab Ind Control Technol, Hangzhou 310027, Zhejiang, Peoples R China
基金
中国国家自然科学基金;
关键词
Bayesian regularization; mixture model; principal component regression (PCR); probabilistic model; soft sensor; PRINCIPAL COMPONENTS; FAULT-DETECTION; SELECTION; BEHAVIOR; SENSORS; NUMBER;
D O I
10.1109/TIE.2014.2385042
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
In this paper, a Bayesian regularization mechanism is provided for automatically determining the number of latent variables in the probabilistic principal component regression (PPCR) model. Different from the unsupervised principal-component-analysis model, the response variable is incorporated for the supervision of selecting latent variables in the PPCR model. By introducing two hyperparameter vectors, the effectiveness of each latent variable can be well measured and controlled. For the mixture form of the PPCR model, a corresponding mixture Bayesian regularization strategy is further developed to control the dimensionality of latent variables. The expectation-maximization algorithm is employed for the parameter learning of both single and mixture Bayesian regularization models. Two probabilistic soft sensors are then developed for the online estimation of key variables in industrial processes, the performances of which are evaluated through two case studies. Compared to the single Bayesian regularization model, the mixture model shows stronger soft sensing abilities in nonlinear and multimode processes.
引用
收藏
页码:4336 / 4343
页数:8
相关论文
共 26 条
[1]  
[Anonymous], 2013, Multivariate statistical process control: process monitoring methods and applications
[2]   Choosing principal components: A new graphical method based on bayesian model selection [J].
Auer, Philipp ;
Gervini, Daniel .
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, 2008, 37 (05) :962-977
[3]   On the optimal selection of principal components in QSPR studies [J].
Besalú, E ;
Vera, L .
JOURNAL OF MATHEMATICAL CHEMISTRY, 2001, 29 (01) :21-34
[4]  
Bishop C., 2006, PATTERN RECOGN, P372
[5]   Behavior Learning in Dwelling Environments With Hidden Markov Models [J].
Bruckner, Dietmar ;
Velik, Rosemarie .
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, 2010, 57 (11) :3653-3660
[6]   Review of Recent Research on Data-Based Process Monitoring [J].
Ge, Zhiqiang ;
Song, Zhihuan ;
Gao, Furong .
INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH, 2013, 52 (10) :3543-3562
[7]   Mixture probabilistic PCR model for soft sensing of multimode processes [J].
Ge, Zhiqiang ;
Gao, Furong ;
Song, Zhihuan .
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, 2011, 105 (01) :91-105
[8]   Principal Component Imagery for the Quality Monitoring of Dynamic Laser Welding Processes [J].
Jaeger, Mark ;
Hamprecht, Fred A. .
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, 2009, 56 (04) :1307-1313
[9]   Just-In-Time Reorganized PCA Integrated with SVDD for Chemical Process Monitoring [J].
Jiang, Qingchao ;
Yan, Xuefeng .
AICHE JOURNAL, 2014, 60 (03) :949-965
[10]   Data-driven Soft Sensors in the process industry [J].
Kadlec, Petr ;
Gabrys, Bogdan ;
Strandt, Sibylle .
COMPUTERS & CHEMICAL ENGINEERING, 2009, 33 (04) :795-814