出行分布的信息熵模型

摘要/Abstract

摘要： 熵最大化方法被广泛应用于交通规划，为交通需求预测模型提供了理论依据.本文使用信息论中的熵概念描述居民出行分布，并根据最大信息熵理论提出了基于原点矩典型特征量约束条件下的出行分布模型，详细阐述了模型中各个参数的实际含义，该模型是一个普适性模型，其具体形式决定于最高阶原点矩阶数的取值，此值对模型精度及参数标定难易程度有重要影响.为确定模型参数，提出了实用的参数标定方法.由于模型中原点矩的最高阶数对模型精度有较大影响，为找到两者之间的关系，利用长春市出行调查数据研究了四种出行方式分别时应于五种原点矩最高阶数的模型拟合情况.因文中所建立的模型是一系列模型，为比较其描述问题的优劣差异，按照最大信息熵原理给出了确定系统分布的熵方法，并运用该方法对前述模型进行了比较.调查数据的验证结果表明，出行分布的信息熵模型可以正确地描述出行分布规律，熵方法可以简单有效地评价不同模型对同一问题描述的优劣差异.本文提出的熵模型和熵方法对城市交通规划具有指导意义.

关键词: 出行分布, 信息熵模型, 最大信息熵原理

Abstract: Entropy maximization approach has been widely used in transportation planning.It provides theoretical basis for a class of forecast models on traffic demand. In this paper，the concept of entropy from information theory is borrowed to describe inhabitant trip distribution and then a trip distribution model is presented under constraints of typical characteristics based on origin moments. The actual meaning of the parameters in the built models is instructed in details. The model is characteristic of generality. The concrete form of the model depends on the order of the highest origin moment which has a great impact on the precision of the model and the labor of parameters calibration. In order to ascertain the parameters in the models，the applied method of calibrating parameters is put forward. Because the number of the most order origin moment has larger effect on the precision of the models，the simulated case of the models of which four trip mode respectively correspond with five most order origin moment is studied by applying the data of trip survey in Changchun city for finding the relation between the two. Because the built models in the paper are a series of models，the entropy method which can ascertain system distribution is educed according to the maximum information entropy theory in order to compare the difference of which the differenmodels describe the same problem. And the above models are compared by using the method. The validated results of survey data show that the information entropy model of trip distribution can correctly describe trip distribution principle and the entropy method can simply and effectively evaluate the difference of which the different models describe the same problem. The entropy models and the entropy method mentioned in the paper have the instructed meaning to traffic planning in city.

Key words: trip distribution, information entropy models, maximum information entropy theory

姚荣涵,王殿海. 出行分布的信息熵模型[J]. 交通运输系统工程与信息, 2005, 5(3): 116-126 .

R. H. YAO,D. H. WANG. Information Entropy Models of Trip Distribution[J]. Journal of Transportation Systems Engineering and Information Technology, 2005, 5(3): 116-126 .

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/abstract/abstract17271.shtml

http://www.tseit.org.cn/CN/Y2005/V5/I3/116

参考文献

[1] 王炜，徐吉谦，及其应用[M].杨涛，李旭宏，等.城市交通规划理论南京:东南大学出版社，1998.
[2] 王炜，徐吉谦，杨涛，李旭宏，等.城市交通规划[M].南京:东南大学出版社，1999.
[3] 沃尔特•赫利，城市系统模型[M].胡运权主译.哈尔滨:哈尔滨工业大学出版社，1987.
[4] 杨兆升.交通运输系统规划— 有关理论与方法[M].北京:人民交通出版社，1998.
[5] Murchland, J. D.. Some Remarks on the Gravity Model of Traffic Distribution and Equivalent Maximizing Formulation[R]. Report LSE-TNT-38, 1966.
[6] Wilson, A. G.，A statistical theory of spatial distribution models [J].Transportation Research,1967, 1:253-2699.
[7] Wilson,A .G ..Entropy in Urban and Regional Modelling. Landon: Pion., 1970,
[8] Wilson,A .G .. Urban and Regional Models in Geography and Planning. London:Wiley,1974.
[9] Potts，R. B.，and Oliver, R. M.. Flows in Transportation Networks. Academic Press，New York and London, 1972.
[10] Willumsen, L.G .. Modelling Transport. John Wiley & Sons，New York,1990.
[11] Hallefjord, AA.，and Jo rnsten, K.. Gravity models with multiple objectives-theory and applications.Transportation Research ,1984,B 20, 19-39.
[12] Li ,Y .，Ziliaskopoulos，T.，and Boyce D.. A Combined Model for Time-dependent Trip Distribution and Traffic Assignment. Paper Presented at the 2002 Annual Meeting of the Transportation Research Board (CD-ROM)，Washington,D .C,2002.
[13] Chang, E.，Ziliaskopoulos, A.，Boyce, D., and Waller, S. T.. So lution Algorithm for the Combined Interregional Commodity Flow and Transportation Network Model with Link Capacity Constraints.Paper Presented at the 2001 Annual Meeting of the Transportation Research Board(CD-ROM) Washington, D .C,2001.
[14] Elmi,A .M .，Badoe, D.A .，and Miller,E .J .. Transferability Analysis of Work-Trip Distribution Models.Paper Presented at the 1999 An nualMeeting of the Transportation Research Board(CD-ROM)，Washington,D .C .，1999.
[15] Wilson,A .G .. The Use of the concept of Entropy in system Modelling. Operations Res. Quart.，1970,21，247-265.
[16] Van Zuylen,H .J .，and Willumsen,L .G ..The Most Likely Trip Matrix Estimated from Traffic Counts.Transportation Research,1980,B 14, 281-293 .
[17] Spiess，H.. A Maximum Likelihood Model for estimating Origin-Destination Matrices Transportation Research,1987，B 21,395-412.
[18] Yu，N.，and Lee, D. H.. An Uncoupled Method for the Equilibrium-Based Linear Path Flow Estimator for Origin-Destination Trip Matrices. Paper Presented at the 2002 Annual Meeting of the Transportation Research Board (CD-ROM)， Washington,D .C .，2002.
[19] List，G. F.，Konieczny, L. A.，and Durnford C. L..A Best Practice Truck Flow Estimation Model for the New York City Region. Paper Presented at the 2002 Annual Meeting of the Transportation Research Board(CD-ROM) Washington, D .C .，2002.
[20] Vovsha，P.. Comparative Analysis of Different Spatial Interaction Models. Paper Presented at the 1999 Annual Meeting of the Transportation Research Board(CD-ROM)，Washington,D .C .，1999.
[21] 颜泽贤.复杂系统演化论[M].北京:人民出版社，1993.
[22] 李如生.平衡和非平衡统计力学[M].北京:清华大学出版社，1995.
[23] 李如生.非平衡态热力学和耗散结构[M].北京:清华大学出版社，1986.
[24] Proakis，J. G..Digital Communications. Beijing: Publishing House of Electronics Industry, 2001.
[25] E.T . Jaynes. Information theory and statistical mechanics[J]. Physical review, 1957, 106(4):620-630.
[26] 哈肯.协同学[M].北京:原子能出版社，1984.
[27] 哈肯.信息与自组织[M].成都:四川教育出版社，1988.
[28] 俞礼军，严海，严宝杰.最大熵原理在交通流统计分布模型中的应用[J].交通运输工程学报，2001, 1(3):91-94.
[29] 哈肯.高等协同学[M].北京:科学出版社，1989.
[30] 曲昭伟，姚荣涵，王殿海.基于最大信息熵原理的居民出行分布模型[J].吉林大学学报[工学版]，2003, 33(2):15-19.
[31] 达庆东，张国伍，姜学峰.交通分布与熵[J].公路交通科技，1999, 16(增1): 36-39.