区域交通流的时空预测与分析.docx

2007 年 6 月J un12007Journal of Highway and Transportation Research and Development文章编号 : 1002Ο0268(2007) 06Ο0092Ο05区域交通流的时空预测与分析韩卫国1 , 王劲峰1 , 高一鸽1 ,2 , 胡建军3(11 中国科学院地理科学与资源研究所 资源与环境信息系统国家重点实验室 , 北京 100101 ;21 中国科学院研究生院 , 北京 100039 ; 31 北京市公安交通管理局科研所 , 北京 100061)摘要 : 论述了短时交通流预测模型的分类 、特点和适用条件 。通过历史交通流量记录运用最优抽样间隔数据分析发现 , 在城市道路网络中 , 路口自身和近邻路口的交通流数据之间存在紧密的时空关系 。利用时空自回归移动平均模型 来建立路口间交通流的时空关联关系 , 用于区域交通流的短时预测和时空分析 , 并详细介绍了该模型的数学描述和建 模过程 。采用长安街及其沿线路口的区域交通流量作为试验数据 , 验证了该模型在交通流的短时预测和时空分析中的 可行性 。该模型在考虑预测值所在位置时间序列的同时 , 也考虑到了空间上相邻位置的时间序列 , 大大提高了短时交 通流预测的准确性 。关键词 : 智能运输系统 ; 交通流时空预测 ; 时空自回归移动平均模型 ; 时空关联关系中图分类号 : U49111 + 4文献标识码 : AFo re c a sting a nd Analysi s of Re gio nal Traffic Flow in Sp a c e a nd TimeHAN WeiΟguo1 , WANG J inΟfeng1 , GAO YiΟge1 ,2 , HU J ianΟjun3(11State Key Laboratory of Resources Environmental Information System , Institute of Geographic Sciences Natural Resources Research , Chinese Academy of Sciences , Beijing 100101 , China ;21 Graduate School of Chinese Academy of Sciences , Beijing 100039 , China ;31Science Research Institute , Beijing Public Security Bureau of Traffic Management , Beijing 100061 , China)Ab stract : Classification , features and applicability of short term traffic flow forecast models are reviewed at first1Based on data analysis of historical traffic volume datasets using the optimized sampling interval method , close spatial and temporal associations exist between traffic flow of the intersection and its neighbors in the urban traffic network1SpaceΟtime autoregressive moving average model is utilized to develop this relationship in space and time between traffic flow of each intersection , and to forecast and explore traffic flow in space and time1Mathematical description and development procedure of the model are also introduced and described specifically1Traffic volumedatasets of the intersections located on the Chang’an Street and its related roads are used as the sample data