Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process

›› 2008, Vol. 16 ›› Issue (5): 746-751.

• SYSTEM ENGINEERING • Previous Articles Next Articles

Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process

FU Yongfeng¹, SU Hongye², ZHANG Ying³, CHU Jian²

1. Modern Education Technology Center, College of Education, Zhejiang University, Hangzhou 310027, China;
2. National Laboratory of Industrial Control Technology, Institute of Cyber Systems and Control, Zhejiang University, Hangzhou 310027, China;
3. BI Center of Competency, IBM China SWG, Shanghai 200021, China

Received:2007-06-22 Revised:2008-05-28 Online:2008-10-28 Published:2008-10-28
Supported by:
the National Natural Science Foundation of China(60421002);priority supported financially by“the NewCentury 151 Talent Project”of Zhejiang Province.

Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process

傅永峰¹, 苏宏业², 张英³, 褚健²

1. Modern Education Technology Center, College of Education, Zhejiang University, Hangzhou 310027, China;
2. National Laboratory of Industrial Control Technology, Institute of Cyber Systems and Control, Zhejiang University, Hangzhou 310027, China;
3. BI Center of Competency, IBM China SWG, Shanghai 200021, China

通讯作者: SU Hongye,E-mail:hysu@iipc.zju.edu.cn
基金资助:
the National Natural Science Foundation of China(60421002);priority supported financially by“the NewCentury 151 Talent Project”of Zhejiang Province.

Abstract

Abstract: To overcome the problem that soft sensor models cannot be updated with the process changes,a soft sensor modeling algorithm based on hybrid fuzzy c-means(FCM)algorithm and incremental support vector machines(ISVM)is proposed.This hybrid algorithm FCMISVM includes three parts:samples clustering based on FCM algorithm,learning algorithm based on ISVM,and heuristic sample displacement method.In the training process,the training samples are first clustered by the FCM algorithm,and then by training each clustering with the SVM algorithm,a sub-model is built to each clustering.In the predicting process,when an incremental sample that represents new operation information is introduced in the model,the fuzzy membership function of the sample to each clustering is first computed by the FCM algorithm.Then,a corresponding SVM sub-model of the clustering with the largest fuzzy membership function is used to predict and perform incremental learning so the model can be updated on-line.An old sample chosen by heuristic sample displacement method is then discarded from the sub-model to control the size of the working set.The proposed method is applied to predict the p-xylene(PX)purity in the adsorption separation process.Simulation results indicate that the proposed method actually increases the model's adaptive abilities to various operation conditions and improves its generalization capability.

Key words: soft sensor, fuzzy c-means, incremental support vector machines, heuristic sample displacement method, p-xylene purity

摘要： To overcome the problem that soft sensor models cannot be updated with the process changes,a soft sensor modeling algorithm based on hybrid fuzzy c-means(FCM)algorithm and incremental support vector machines(ISVM)is proposed.This hybrid algorithm FCMISVM includes three parts:samples clustering based on FCM algorithm,learning algorithm based on ISVM,and heuristic sample displacement method.In the training process,the training samples are first clustered by the FCM algorithm,and then by training each clustering with the SVM algorithm,a sub-model is built to each clustering.In the predicting process,when an incremental sample that represents new operation information is introduced in the model,the fuzzy membership function of the sample to each clustering is first computed by the FCM algorithm.Then,a corresponding SVM sub-model of the clustering with the largest fuzzy membership function is used to predict and perform incremental learning so the model can be updated on-line.An old sample chosen by heuristic sample displacement method is then discarded from the sub-model to control the size of the working set.The proposed method is applied to predict the p-xylene(PX)purity in the adsorption separation process.Simulation results indicate that the proposed method actually increases the model's adaptive abilities to various operation conditions and improves its generalization capability.

关键词: soft sensor, fuzzy c-means, incremental support vector machines, heuristic sample displacement method, p-xylene purity

FU Yongfeng, SU Hongye, ZHANG Ying, CHU Jian. Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process[J]. , 2008, 16(5): 746-751.

傅永峰, 苏宏业, 张英, 褚健. Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process[J]. , 2008, 16(5): 746-751.

References

1 McAvoy, T.J., “Contemplative stance for chemical process control”, Automatica, 28 (2), 441-442 (1992).
2 Zhong, W., Liu, A.L., Yu, J.S., “Studies on building soft-sensing models of multivariable systems”, Control and Decision, 15 (2), 209-212 (2000).
3 Liu, R.L., “Some studies on soft sensor technology and their applications to industry process”, Ph.D. Thesis, Zhejiang Univ., China (2004). (in Chinese) .
4 Xiao, R., Wang, J.C., Sun, Z.X., Zhang, F.Y., “An SVM incremental learning algorithm α-ISVM”, Journal of Software, 12 (12), 1818-1824 (2001).
5 Vapnik, V.N., Statistical Learning Theory (translated by Xu, J.H., Zhang, X.G.), Publishing House of Electronics Industry, Beijing (2004).
6 Yan, W.W., Shao, H.H., “Application of support vector machines and least squares support vector machines to heart disease diagnoses”, Control and Decision, 18 (3), 358-360 (2003).
7 Dunn, J.C., “A fuzzy relative of the ISODATA process and its use in detecting compact well separated cluster”, Journal of Cybernet, 3, 32-57 (1974).
8 Bezdek, J.C., “A convergence theorem for the fuzzy ISODATA clustering algorithm”, IEEE Trans Pattern Anal March Intel, 2 (1), 1-8 (1980).
9 Burges, C.J.C., “A tutorial on support vector machines for pattern recognition”, Data Mining and Knowledge Discovery, 2 (2), 1-47 (1998).
10 Chang, C.C., Lin, C.J., “Training nu-support vector regression:theory and algorithms”, Neural Computation, 14 (2), 1959-1977 (2002).
11 Naqa, I.E., Yang, Y., “Relevance feedback based on incremental learning for mammogram retrieval”, Image Processing, 4 (2), 7-11 (2003).
12 Cauwenberghs, G., Poggio, T., “Incremental and decremental support vector machine learning”, Advances in Neural Information Processing Systems, (NIPS' 2000), 13, 409-415, MIT Press, MA (2001).
13 Tax, D.M.J., Duin, R.P.W., “Data domain description using support vectors”, In:Proc. of 8th European Symposium on Artificial Neural Networks, D-Facto Public., Bruges, Belgium, 251-256 (1999).
14 Zhang, Y., “A study on process industrial data mining based on support vector machines”, Ph. D. Thesis, Zhejiang Univ., China (2005). (in Chinese) .
15 Pavone, D., Hotier, G., “System approach modeling applied to the eluxyl process”, Oil & Gas Science and Technology-rev, 55 (4), 437-446 (2000).
16 Minceva, M., Rodrigues, A.E., “Modeling and simulation of a simulated moving bed for the separation of p-xylene”, Ind. Eng. Chem. Res., 41, 3454-3461 (2002).

[1]	Gongpeng Wu, Yan He. Investigation on gas–solid flow regimes in a novel multistage fluidized bed [J]. Chinese Journal of Chemical Engineering, 2022, 46(6): 21-30.
[2]	Zheng Wang, Cheng Shao, Li Zhu. Soft-sensing modeling and intelligent optimal control strategy for distillation yield rate of atmospheric distillation oil refining process [J]. Chinese Journal of Chemical Engineering, 2019, 27(5): 1113-1124.
[3]	Kaixun He, Kai Wang, Yayun Yan. Active training sample selection and updating strategy for near-infrared model with an industrial application [J]. Chinese Journal of Chemical Engineering, 2019, 27(11): 2749-2758.
[4]	Honggui Han, Shuguang Zhu, Junfei Qiao, Min Guo. Data-driven intelligent monitoring system for key variables in wastewater treatment process [J]. Chin.J.Chem.Eng., 2018, 26(10): 2093-2101.
[5]	Pengfei Cao, Xionglin Luo, Xiaohong Song. Modeling and identification for soft sensor systems based on the separation of multi-dynamic and static characteristics [J]. Chin.J.Chem.Eng., 2018, 26(1): 137-143.
[6]	Pengfei Cao, Xionglin Luo, Xiaohong Song. Feasibility analysis and online adjustment of constraints in model predictive control integrated with soft sensor [J]. , 2017, 25(9): 1230-1237.
[7]	Huan Min, Xionglin Luo. Calibration of soft sensor by using Just-in-time modeling and AdaBoost learning method [J]. , 2016, 24(8): 1038-1046.
[8]	Wentao Cang, Huizhong Yang. Estimation of catalytic activity using an unscented Kalman filtering in condensation reaction [J]. Chin.J.Chem.Eng., 2015, 23(12): 1965-1969.
[9]	Weiming Shao, Xuemin Tian, PingWang. Supervised local and non-local structure preserving projections with application to just-in-time learning for adaptive soft sensor [J]. Chin.J.Chem.Eng., 2015, 23(12): 1925-1934.
[10]	Yanlin He, Yuan Xu, Zhiqiang Geng, Qunxiong Zhu . Soft sensor of chemical processes with large numbers of input parameters using auto-associative hierarchical neural network [J]. Chin.J.Chem.Eng., 2015, 23(1): 138-145.
[11]	Ouguan Xu, Yongfeng Fu, Hongye Su, Lijuan Li . A Selective Moving Window Partial Least Squares Method and Its Application in Process Modeling [J]. , 2014, 22(7): 799-804.
[12]	CAO Pengfei, LUO Xionglin. Soft Sensor Model Derived from Wiener Model Structure：Modeling and Identification [J]. Chin.J.Chem.Eng., 2014, 22(5): 538-548.
[13]	LÜ Ye, YANG Huizhong. A Multi-model Approach for Soft Sensor Development Based on Feature Extraction Using Weighted Kernel Fisher Criterion [J]. Chin.J.Chem.Eng., 2014, 22(2): 146-152.
[14]	XU Feng, WANG Yeye, LUO Xionglin. Soft Sensor for Inputs and Parameters Using Nonlinear Singular State Observer in Chemical Processes [J]. Chin.J.Chem.Eng., 2013, 21(9): 1038-1047.
[15]	YAN Xingdi, YANG Wen, MA Hehe, SHI Hongbo. Soft Sensor for Ammonia Concentration at the Ammonia Converter Outlet Based on an Improved Group Search Optimization and BP Neural Network^* [J]. Chin.J.Chem.Eng., 2012, 20(6): 1184-1190.

Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process

Adaptive Soft-sensor Modeling Algorithm Based on FCMISVM and Its Application in PX Adsorption Separation Process

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics

Comments