Please wait a minute...
浙江大学学报(工学版)
J4  2013, Vol. 47 Issue (10): 1741-1746    DOI: 10.3785/j.issn.1008-973X.2013.10.006
    
Multi-model predictive control based on AP-LSSVM
LI Li-juan1, XIONG Lu1, LIU Jun1, XU Ou-guan2
1. School of Automation and Electrical Engineering, Nanjing University of Technology, Nanjing 210009, China; 2. Zhijiang College, Zhejiang University of Technology, Hangzhou 310024, China
Download:   PDF(0KB) HTML
Export: BibTeX | EndNote (RIS)      

Abstract  

A new multi-model predictive control algorithm, integrating affinity propagation (AP) clustering and least squares support vector machines (LS-SVM) modeling technology with particle swarm optimization (PSO) algorithm, was presented in order to enhance the control property of a nonlinear system. Samples were clustered into several classes by AP algorithm and the training data of every class were obtained. Then corresponding sub-models were constructed as the predictive model by LS-SVM whose parameters were optimized through grid-search and cross-validation method. In rolling optimization stage, the sub-model was selected by computing the Euclidean distances between current control variable and each center of clustering, and corresponding predictive outputs were generated. Then the optimization problem was constructed according to the predictive outputs and reference trajectory. The optimal control values were obtained by the rolling PSO optimization algorithm. The presented algorithm was applied in the simulation of aromatics isomerization process. The presented algorithm, single model based predictive control algorithm and multi-models predictive control algorithm based on k-means and BP neural network were applied, respectively. Results show the better performance of the proposed multi-model predictive control algorithm.

Published: 01 October 2013
CLC:  TP 13  
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
Cite this article:

LI Li-juan, XIONG Lu, LIU Jun, XU Ou-guan. Multi-model predictive control based on AP-LSSVM. J4, 2013, 47(10): 1741-1746.

URL:

http://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2013.10.006     OR     http://www.zjujournals.com/eng/Y2013/V47/I10/1741


基于AP-LSSVM的多模型预测控制

针对一类非线性系统的控制问题,结合基于仿射传播聚类的最小二乘支持向量机（LS-SVM）多模型建模算法与PSO优化算法,提出新的多模型预测控制算法.采用仿射传播聚类算法对历史样本数据进行聚类,得到各个类的训练样本数据|利用LS-SVM对各个类分别建立子模型,采用网格搜索和交叉验证为各子模型找到合适的模型参数,将所建立的子模型作为预测控制算法的预测模型.在滚动优化时,计算当前控制量与各聚类中心的欧氏距离,选择相应的子模型计算未来时刻模型的预测输出,计算得到参考轨迹.建立优化问题的目标函数,采用PSO算法优化求解得到系统的最优控制量作用于对象.将提出的算法在某芳烃异构化过程中进行仿真试验，分别采用提出的算法以及单模型预测控制算法、基于k均值和BP神经网络的多模型预测控制算法进行仿真.结果表明,采用提出的多模型预测控制算法可以获得更好的控制性能.

［1］ PAOLO L, BRUNO M, CLAUDIO A. Modeling and predictive control of a new injection system for compressed natural gas engines ［J］. Control Engineering Practice, 2008, 16(10): 1216-1230.

［2］ 席裕庚,庚晓军,陈虹.预测控制性能研究新进展［J］.控制理论与应用,2000, 17(4): 469-485.

XI Yu-geng, GENG Xiao-jun, CHEN Hong. New progress of predictive control performance study ［J］. Control Theory and Applications, 2000, 17(4): 469-485.

［3］ LI Qiang, QU Bao-cun, GE Zhi-qiang. Study of fuzzy generalized predictive control algorithm on nonlinear systems ［C］∥Proceedings of the 1st International Conference on Innovative Computing, Information and Control. Beijing: IEEE, 2006: 437-440.

［4］ 刘雪芹,刘晓华.基于多神经元模型的非线性系统预测控制［J］.控制工程,2005, 12(supplement): 128-130.

LIU Xue-qin, LIU Xiao-hua. Nonlinear predictive control system based on multi-neuron model ［J］. Control Engineering, 2005, 12(增刊): 128-130.

［5］ ZHONG Wei-min, PI Dao-ying, SUN You-xian. Study on SVM based model predictive control ［C］∥Proceedings of the 5th Word Congress on Intelligent Control and Automation (in Chinese). Hangzhou: IEEE, 2004: 607-610.

［6］ 杨红,罗飞,许玉格,等.基于混沌优化的LS-SVM非线性预测控制方法［J］.计算机工程与应用,2010, 46(5): 229-232.

YANG Hong, LUO Fei, XU Yu-ge, et al. LS-SVM nonlinear control method based on chaos optimization algorithm ［J］. Computer Control and Application, 2010, 46(5): 229-232.

［7］ 穆朝絮,张瑞民,孙长银.基于粒子群优化的非线性系统最小二乘支持向量机预测控制方法［J］.控制理论与应用, 2010, 27(2): 164-168.

MU Zhao-xu, ZHANG Rui-min, SUN Chang-yin. LS-SVM nonlinear control method based on particle swarm optimization ［J］. Control Theory and Applications, 2010, 27(2): 164-168.

［8］ 李丽娟,刘君.芳烃异构化过程的多模型建模［J］.化工学报, 2011, 62(8): 2546-2552.

LI li-juan, LIU Jun. Multi-modeling of aromatics isomerization process ［J］. CIESC Journal, 2011, 62(8): 2546-2552.

［9］ 周立芳,张赫男.基于聚类多模型建模的测控制［J］.化工学报, 2008, 59(10): 2350-2354.

ZHOU Li-fang, ZHANG He-nan. Multi-model predictive control based on cluster multi-modeling ［J］. CIESC Journal, 2008, 59(10): 2350-2354.

［10］ FREY B J, DUECK D. Clustering by passing messages between data points ［J］. Science, 2007, 315(5814): 972976.

［11］ VAPNIK V. The nature of statistical learning theory ［M］. New York: Springer, 1995.

［12］ SUYKENS J A K, VANDEWALLE J. Multiclass least squares support vector machines ［C］∥International Joint Conference on Neural Networks. Washington: ［s. n.］, 1999.

［13］ KENNEDY J, EBERHART R C. Particle swarm optimization ［C］∥Proceeding of IEEE International Conference on Neural Networks. Piscataway: IEEE, 1995: 1942-1948.

［14］ 徐欧官.芳烃联合装置芳烃转化过程建模与应用研究［D］.杭州：浙江大学, 2007.

XU Ou-guan. Aromatics integrated unit and aromatics conversion process modeling and application ［D］. Hangzhou: Zhejiang University, 2007.
[1] LOU Ke, QI Bin, MU Wen-ying, CUI Bao-tong. Flocking control of multiple agents based on feedback control strategy[J]. J4, 2013, 47(10): 1758-1763.
[2] MAO Wei-jie, ZHANG Yuan-yuan. Stability analysis for neutral systems with interval time-varying delays[J]. J4, 2012, 46(5): 848-852.
[3] WANG Li-juan, ZHANG Hui. On parameter identifiability of linear multivariable systems under
communication access constraints[J]. J4, 2011, 45(12): 2103-2108.
[4] NI He, XIAO Hang, CHENG Gang, SUN Feng-rui. Optimization of load controller for steam power plant using
evolutionary algorithm and chaos[J]. J4, 2011, 45(12): 2202-2207.
[5] ZHAO Li-li, LI Ping, LI Xiu-liang. Design of adaptive observer with forgetting factor for linear system[J]. J4, 2011, 45(10): 1704-1709.
[6] WANG Dun-Hong, XUE An-Ke. Output feedback control for timedelay system with
quantized measurement[J]. J4, 2010, 44(7): 1418-1422.
[7] BO Hai-Feng, LV Yong-Song. Fuzzy neural network control method with compensation
for time-delay system[J]. J4, 2010, 44(7): 1343-1347.
[8] JU Hong-Bei, FANG Shan-Hua, LU Ren-Quan, XUE An-Ke. Design of quantized state estimator for dualstage control system[J]. J4, 2010, 44(7): 1428-1432.
[9] XIE Bin, ZHANG Kai-Dan, BO Hua-Dong, et al. Hospitability map building and ground target tracking with UAV under unknown environments[J]. J4, 2010, 44(1): 118-123.