Please wait a minute...
浙江大学学报(工学版)
J4  2009, Vol. 43 Issue (5): 817-821    DOI: 10.3785/j.issn.1008-973X.2009.05.006
自动化技术、计算机技术     
基于交叉格莱姆矩阵的最小信息损失模型降阶方法
付金宝,章辉,孙优贤
（浙江大学 工业控制技术国家重点实验室,浙江 杭州 310027）
Model reduction by minimizing information loss based on cross-Gramian matrix
 FU Jin-Bao, ZHANG Hui, SUN You-Xian
(State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou 310027, China)
 全文: PDF(606 KB)   HTML
摘要：

运用信息论的原理和方法研究以状态空间形式描述的线性定常系统的模型降阶问题.应用线性定常系统及其对偶系统的稳态状态信息熵规范了系统能控性信息和能观性信息的定义形式.基于交叉格莱姆矩阵的系统信息属性,定义了交叉格莱姆信息.通过分析系统稳态状态信息熵的信息描述形式,揭示了交叉格莱姆信息的状态物理含义.在基于最小信息损失的模型降阶过程中以交叉格莱姆信息损失最小为目标,提出了新的模型降阶方法——CGMIL方法.理论分析和仿真结果表明,交叉格莱姆信息是包含系统能控性信息和能观性信息的综合信息描述形式,CGMIL方法与基于最小信息损失的模型降阶方法相比能够获得更好的降阶性能.
Abstract:

The model reduction of the continuous linear time-invariant (LTI) system described by the form of state space was studied by adopting the information theoretic  principle and method. The definitions of controllability information and observability information were standardized by using the information entropy of steady states of the LTI  system and its dual system. Cross-Gramian information (CGI) was defined based on the information theoretic properties of the systems cross-Gramian matrix. By analyzing the  information descriptions of system states, the physical meaning of CGI was clarified from the view of entropic points. By taking minimizing the loss of CGI as the performance  index of model reduction, an improved model reduction method by minimizing the CGI loss, denoted by CGMIL, was developed. The theoretical analysis and simulation indicated that  CGI is a comprehensive information description including both the controllability information and the observability information; and CGMIL algorithm is much better than the MIL  algorithm in the performance of model reduction．
出版日期: 2009-11-18
:  TP14  
基金资助:

国家自然科学基金资助项目（60674028，60736021，60674071）；国家“863”高技术研究发展计划资助项目（2006AA04Z184）.
通讯作者: 章辉，男，副教授.     E-mail: zhanghui@iipc.zju.edu.cn
作者简介: 付金宝(1982-)，男，黑龙江双城人，硕士生，从事线性随机系统模型降阶研究.
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章　　

引用本文:

付金宝, 章辉, 孙优贤. 基于交叉格莱姆矩阵的最小信息损失模型降阶方法[J]. J4, 2009, 43(5): 817-821.

FU Jin-Bao, ZHANG Hui, SUN You-Xian. Model reduction by minimizing information loss based on cross-Gramian matrix. J4, 2009, 43(5): 817-821.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2009.05.006    或    http://www.zjujournals.com/eng/CN/Y2009/V43/I5/817

［1］ LELAND R. Reduced-order models and controllers for continuous time stochastic systems: an information theory approach ［J］. IEEE Transactions on Automatic Control, 1999,  44(9): 17141719．
［2］ ZHANG Hui, SUN You-xian. Information theoretic methods for stochastic model reduction based on state projection ［C］∥ 2005 American Control Conference. Oregon: ［s.n.］,  2005: 25962601．
［3］ 章辉. 控制系统中的信息描述与方法［D］. 杭州：浙江大学, 2003: 5182．
ZHANG Hui. Information descriptions and approaches in control systems ［D］. Hangzhou: Zhejiang University, 2003: 5182．
［4］ 曹清,章辉,孙优贤. 改进的用于模型降阶的最小信息损失方法［J］. 信息与控制, 2005, 34(4): 423428．
CAO Qing, ZHANG Hui, SUN You-xian. A revised minimum information loss method for model reduction ［J］. Information and Control, 2005, 34(4): 423428．
［5］ FERNADO K V, NICHOLSON H. On the structure of balanced and other principal representations of SISO systems ［J］. IEEE Transactions on Automatic Control, 1983, AC-28(2): 228231．
［6］ FERNADO K V, NICHOLSON H. Minimality of SISO linear systems ［J］. Proceedings of IEEE, 1982, 70(10): 12411242．
［7］ WGS. Benchmarks examples for model reduction of linear time invariant dynamical systems ［EB/OL］. ［2007-10］. http:∥www.icm.tu-bs.de/NICONET/index.html．
［8］ WGS. A collection of benchmarks examples for model reduction of linear time invariant dynamical systems ［EB/OL］. ［2007-10］. http:∥www.icm.tu-bs.de/NICONET/index.html
No related articles found!