Soft-Sensing Method with Online Correction Based on Semi-Supervised Learning

doi:10.1007/s12204-015-1606-y

摘要/Abstract

摘要： Soft sensing has been widely used in chemical industry to build an online monitor of the variables which are unmeasurable online or measurable online but with a high cost. One inherent difficulty is insufficiency of the training samples because the labeled data are limited. Besides, the traditional soft-sensing structure has no online correction mechanism. The forecasting result may be incorrect if the working condition is changed. In this work, a semi-supervised learning (SSL) method is proposed to build the soft-sensing model by use of the unlabeled data. Meanwhile, an online correction mechanism is proposed to establish a soft-sensing approach. The mechanism estimates the input variables at each step by a prediction model and calibrates the output variables by a compensation model. The experimental results show that the proposed method has better prediction accuracy and generalization ability than other approaches.

关键词: soft-sensing, semi-supervised learning (SSL), online correction, neural network

Abstract: Soft sensing has been widely used in chemical industry to build an online monitor of the variables which are unmeasurable online or measurable online but with a high cost. One inherent difficulty is insufficiency of the training samples because the labeled data are limited. Besides, the traditional soft-sensing structure has no online correction mechanism. The forecasting result may be incorrect if the working condition is changed. In this work, a semi-supervised learning (SSL) method is proposed to build the soft-sensing model by use of the unlabeled data. Meanwhile, an online correction mechanism is proposed to establish a soft-sensing approach. The mechanism estimates the input variables at each step by a prediction model and calibrates the output variables by a compensation model. The experimental results show that the proposed method has better prediction accuracy and generalization ability than other approaches.

Key words: soft-sensing, semi-supervised learning (SSL), online correction, neural network

中图分类号:

TP 274.2

TANG Qi-feng (汤奇峰), LI De-wei* (李德伟), XI Yu-geng (席裕庚). Soft-Sensing Method with Online Correction Based on Semi-Supervised Learning[J]. 上海交通大学学报（英文版）, 2015, 20(2): 171-176.

TANG Qi-feng (汤奇峰), LI De-wei* (李德伟), XI Yu-geng (席裕庚). Soft-Sensing Method with Online Correction Based on Semi-Supervised Learning[J]. Journal of shanghai Jiaotong University (Science), 2015, 20(2): 171-176.

参考文献 17

[1]	Gonzaga J C B, Meleiro L A C, Kiang C, et al.ANN-based soft-sensor for real-time process monitoring and control of an industrial polymerization process[J]. Computer Chemical Engineering, 2009, 34(1): 43-49.
[2]	Topcu I B, Saridemir M. Prediction of rubberized mortar properties using artificial neural network and fuzzy logic [J]. Journal of Materials Processing Technology,2008, 199(1-3): 108-118.
[3]	Vijayabaskar V, Gupta R, Chakrabarti P P, et al. Prediction of properties of rubber by using artificial neural networks [J]. Journal of Applied Polymer Science, 2006, 100(3): 2227-2237.
[4]	Facco P, Doplicher F, Bezzo F, et al. Moving average PLS soft sensor for online product quality estimation in an industrial batch polymerization process[J]. Journal of Process Control, 2009, 19(3): 520-529.
[5]	Liu X Q, Kruger U, Littler T, et al. Moving window kernel PCA for adaptive monitoring of nonlinear processes [J]. Chemometrics and Intelligent Laboratory System, 2009, 96(2): 132-143.
[6]	Kim K, Lee J M, Lee I B. A novel multivariate regression approach based on kernel partial least squares with orthogonal signal correction [J]. Chemometrics and Intelligent Laboratory System, 2009, 79(1-2): 22-30.
[7]	Su Z G, Wang P H, Shen J, et al. Multi-model strategy based evidential soft sensor model for predicting evaluation of variables with uncertainty [J]. Applied Soft Computing, 2011, 11(2): 2595-2610.
[8]	Zhong W, Yu J S. MIMO soft sensors for estimating product quality with online correction [J]. Chemical Engineering Research and Design, 2000, 78(4): 612-620.
[9]	Peng Xiao-qi, Sun Yuan, Tang Ying. Performance monitoring and assessment of a soft-sensor and its adaptive correction [J]. Journal of Chemical Industry and Engineering, 2012, 63(5): 1474-1483 (in Chinese).
[10]	Tang Q F, Li D W, Xi Y G, et al. Soft-sensing design based on semiclosed-loop framework [J]. Chinese Journal of Chemical Engineering, 2012, 20(6): 1213-1218.
[11]	Nigam K, Mccallum A K, Thrun S, et al. Text classification from labeled and unlabeled documents using EM [J]. Machine Learning, 2000, 39(2-3): 103-134.
[12]	Lam H K, Ling S H, Tam P K S, et al. Tuning of the structure and parameters of neural networks using an improved genetic algorithm [J]. IEEE Transactions on Neural Network, 2003, 14(1): 79-88.
[13]	Eriksson M, Golriz M R. Radiation heat transfer in circulating fluidized bed combustors [J]. International Journal of Thermal Sciences, 2005, 44(4): 399-409.
[14]	Guedea I, Bolea I, Lupi′a?nez C, et al. Control system for an oxy-fuel combustion fluidized bed with flue gas recirculation [J]. Energy Procedia, 2011, 4: 972-979.
[15]	Tang Q F, Zhao L, Qi R B, et al. Tuning the structure and parameters of a neural network by using cooperative quantum particle swarm algorithm [J]. Measuring Technology and Mechatronics Automation, 2011,48: 1328-1332.
[16]	Tang Qing-feng. The cooperative quantum-particle swarm algorithm and its application in the energy utilization optimization of the steam network [D]. Shanghai:East China University of Science & Technology,2011 (in Chiense).
[17]	Yeh T T, Espina P I, Osella S A. An intelligent ultrasonic flow meter for improved flow measurement and flow calibration facility [C]//Instrumentation and Measurement Technology Conference. Budapest, Hungary:IEEE, 2001: 1741-1746.