基于自适应迭代学习控制的MPC系统经济性能设计

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

清华大学辅仁网/2017-07-07

王振雷¹, 刘学彦¹, 王昕²

1. 华东理工大学化工过程先进控制和优化技术教育部重点实验室, 上海 200237;
2. 上海交通大学电工与电子技术中心, 上海 200240

Economic performance design based on adaptive iterative learning control of MPC systems

WANG Zhenlei¹, LIU Xueyan¹, WANG Xin²

1. Key Laboratory of Advanced Control and Optimization for Chemical Processes of the Ministry of Education, East China University of Science and Technology, Shanghai 200237, China;
2. Center of Electrical and Electronic Technology, Shanghai Jiao Tong University, Shanghai 200240, China

摘要:

输出: BibTeX | EndNote (RIS)

摘要针对模型预测控制（model predictive control，MPC）系统经济性能设计问题，结合自适应迭代学习控制的设计思想，提出了一种自适应步长迭代学习控制（adaptive step iterative learning control，ASILC）策略。该策略将系统变量方差与控制器参数之间的关系近似成离散的线性区间组合，并借助上一步迭代的过程信息，自适应地更新迭代步长，逐步使系统的经济性能达到最优。将该方法应用于乙烯裂解炉控制系统中，仿真结果表明：与迭代学习控制方法相比，ASILC能更快地收敛到最优工作点附近，得到最优经济性能下的控制器参数λ，经过7次优化迭代后经济性能目标值提高了28.92%。

关键词 ：自适应迭代学习控制,经济性能设计,模型预测控制(MPC),迭代步长,乙烯裂解炉

Abstract：An adaptive step iterative learning control (ASILC) strategy was developed for model predictive control (MPC) system economic performance design. The strategy treats the functional relationship between the variable variances and the controller parameters as a combination of discrete linear intervals and uses process information in the last iteration to adaptively update the iteration step. This optimizes the economic performance step by step. The method is used to design an ethylene cracking furnace control system. Simulations show that ASILC converges to the optimal operating point faster than iterative learning control (ILC) and obtains the controller parameter λ for the optimal economic performance. After seven optimizations and iterations, the economic performance target was improved 28.92%.

Key words：adaptive iterative learning control economic performance design model predictive control (MPC)iterative step ethylene cracking furnace

收稿日期: 2015-08-25 出版日期: 2016-09-22

ZTFLH:

TP273

引用本文:

王振雷, 刘学彦, 王昕. 基于自适应迭代学习控制的MPC系统经济性能设计[J]. 清华大学学报（自然科学版）, 2016, 56(9): 1016-1024.
WANG Zhenlei, LIU Xueyan, WANG Xin. Economic performance design based on adaptive iterative learning control of MPC systems. Journal of Tsinghua University(Science and Technology), 2016, 56(9): 1016-1024.

链接本文:

http://jst.tsinghuajournals.com/CN/10.16511/j.cnki.qhdxxb.2016.24.024或 http://jst.tsinghuajournals.com/CN/Y2016/V56/I9/1016

图表:

图1 MPC系统经济性能设计的双层架构

图2 单输入单输出系统的LQG 权衡曲线

图3 离散的线性函数区间组合

图4 带约束最优规划问题示意图(输入为积极约束)

表1 迭代方向的选择策略

图6 ASILC策略的控制流程图

图6 乙烯裂解炉原理简图

图7 底层MPC系统的输入输出数据变化

图8 迭代过程中整定参数和经济性能随迭代步数的变化趋势

参考文献:

[1] Harris T J. Assessment of control loop performance [J]. Canadian Journal of Chemical Engineering, 1989, 67(5): 856-861.
[2] HUANG Biao, Shah S L, Kwok E K. Good, bad or optimal? Performance assessment of multivariable processes [J]. Automatic, 1997, 33 (6): 1175-1183.
[3] HUANG Biao, Shah S L. Performance Assessment of Control Loops: Theory and Applications [M]. London, UK: Springer, 1999.
[4] HUANG Biao. A pragmatic approach towards assessment of control loop performance [J]. International Journal of Adaptive Control and Signal Processing, 2003, 17(729): 589-608.
[5] Grimble M J. Controller performance benchmarking and tuning using generalized minimum variance control [J]. Automatic, 2002, 38(12): 2111-2119.
[6] Bauer M, Craig I K. Economic assessment of advanced process control—A survey and framework [J]. Journal of Process Control, 2008, 18(1): 2-18.
[7] Martin G D, Turpin L E, Cline R P. Estimating control function benefits [J]. Hydrocarbon Processing, 1991, 70(6): 68-73.
[8] Latour P R. Quantify quality control’s intangible benefits [J]. Hydrocarbon Processing, 1992, 71(5): 61-65.
[9] Bao J, Mclellan P, Forbes J. Towards a systematic approach for control system benefits analysis [C]//Proceedings of the Control Systems, Preprints, Conference. Norcross, USA: Technical Association of the Pulp and Paper Industry Press, 2000: 334-340.
[10] 赵超. 过程控制系统经济性能评估算法的研究 [D]. 杭州: 浙江大学, 2009. ZHAO Chao. Economic Performance Assessment Algorithm Research of Process Control System [D]. Hangzhou: Zhejiang University, 2009. (in Chinese)
[11] Sorensen R C, Cutler C R. LP integrates economics into dynamic matrix control [J]. Hydrocarbon Processing, 1998, 77(9): 57-65.
[12] Krishman A, Kosanovich K, De Witt M, et al. Robust model predictive control of an industrial solid phase polymerizer [C]//Proceedings of the American control Conference. Philadelphia, USA: American Automatic Control Council, 1998.
[13] XU Fangwei, HUANG Biao, Akande S. Performance Assessment of model predictive control for variability and constraint tuning [J]. Industrial and Engineering Chemistry Research, 2007, 46(4): 1208-1219.
[14] ZHAO Chao, ZHAO Yu, SU Hongye, et al. Economic performance assessment of advanced process control with LQG benchmarking [J]. Journal of Process Control, 2009, 19(4): 557-569.
[15] Moore K L. Iterative Learning Control for Deterministic Systems [M]. London, UK: Springer-Verlag, 1993.
[16] Ahn H S, CHEN Yangquan, Moore K L. Iterative learning control: Brief survey and categorization [J] IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 2007, 37(6): 1099-1121.
[17] CAI Xing, SUN Pei, CHEN Junhui, et al. ILC strategy for progress improvement of economic performance in industrial model predictive control systems [J]. Journal of Process Control, 2014, 24(12): 373-389.
[18] 杜志国, 曾清泉, 陈硕. 管式裂解炉辐射区裂解管结焦模型进展 [J]. 石油化工, 2003, 32(4): 348-352. DU Zhiguo, ZENG Qingquan, CHEN Shuo. Progress in coking model of tube in radiation section of cracking furnace [J]. Petrochemical Technology, 2003, 32(4): 348-352. (in Chinese)

[1]	王振雷, 毛福兴, 王昕. 基于二阶段自适应多模型的聚合釜温度控制[J]. 清华大学学报（自然科学版）, 2016, 56(7): 707-716.
[2]	王振雷, 陈登乾, 王昕. 基于多模型切换策略的精馏塔最优控制[J]. 清华大学学报（自然科学版）, 2016, 56(4): 430-436.
[3]	施磊, 江永亨, 王凌, 黄德先. 一种求解炼油厂连续时间调度模型的Lagrange分解算法[J]. 清华大学学报（自然科学版）, 2016, 56(4): 437-447.
[4]	黄涛, 杨开明, 朱煜, 蒋毅, 胡楚雄. 3自由度硅片台解耦时变参数在线递推估计[J]. 清华大学学报（自然科学版）, 2016, 56(2): 185-191.