基于时空分析的短时交通流量预测模型*

南京大学学报(自然科学版) ›› 2010, Vol. 46 ›› Issue (5): 552–560.

基于时空分析的短时交通流量预测模型*

夏英 1, 2 ** , 梁中军 2 , 王国胤 2

出版日期:2015-04-02 发布日期:2015-04-02
作者简介: (1. 西南交通大学信息科学与技术学院, 成都, 610031;
2. 重庆邮电大学计算机科学与技术学院, 重庆, 400065)
基金资助:
国家自然科学基金( No. 60773113), 重庆市科委科技项目( CSTCA2009CB2015) , 重庆市计算机网络与通信技术
重点实验室开放基金( CY- CNCL- 2009- 01)

Research of short-term traffic flow forecasting model based on spatio-temporal analysis

Xia Ying 1, 2 , Liang Zhong Jun 2 , Wang Guo Yin 2

Online:2015-04-02 Published:2015-04-02
About author: (1. School of Information Science and T echnology, Southwest Jiaotong University, Chengdu, 610031, China;
2. School of Computer Science and Technology, Chongqing University of Posts and
T elecommunications, Chongqing, 400065, China)

摘要/Abstract

摘要： 根据交通流的时空关联性和非线性, 提出一种基于时空分析的短时交通流量预测模型. 在相关系数的基础上扩展时空语义, 提出时空相关分析算法, 并以支持向量机为预测工具进行预测. 弥补现
有模型在预测因子选取方面的不足, 提高预测精度并避免预测的人为主观性. 实验结果表明了算法和模型的有效性.

Abstract: Short-term traffic flow forecasting is one of the important issues for intelligent transportation systems ( ITS) . T here are various methods have been developed to forecast traffic flow so far, but most of the forecasting
models are constructed according to analysis of the historical and current traffic flow series in selected section or crossing, without considering the spatial information of related traffic network. Actually, there are many spatio-
temporal information such as spatial connectivity of the traffic network and time latency of traffic flow series, these characteristics can be used to improve the efficiency of forecasting. A short-term traffic flow forecasting method
based on spatio -temporal analysis is proposed here. Firstly, spatio-temporal correlation coefficient is defined to reflect the relationship of different traffic flow series, a quick calculation method of spatio-temporal correlation
coefficient is proposed after analyzing the corresponding properties. Secondly, a spatio-temporal analysis algorithm based on spatio-temporal correlation coefficient matrix is proposed to choose the proper predictor. At last, a
forecasting model is built based on support vector machine due to the nonlinear characteristics of traffic flow. To evaluate the proposed method and forecasting model, temporal correlation analysis, spatial correlation analysis and
principal component analysis are utilized and tested on synthetic traffic datasets named DynaCHINA. Experimental results show that the proper spatio?temporal characteristics of the traffic flows can be selected as input variables of
the forecasting model effectively and the forecasting accuracy is improved.

夏英 1, 2 ** , 梁中军 2 , 王国胤 2
. 基于时空分析的短时交通流量预测模型*
[J]. 南京大学学报(自然科学版), 2010, 46(5): 552–560.

Xia Ying 1, 2 , Liang Zhong Jun 2 , Wang Guo Yin 2 . Research of short-term traffic flow forecasting model based on spatio-temporal analysis

[J]. Journal of Nanjing University(Natural Sciences), 2010, 46(5): 552–560.

参考文献

[ 1 ] Gao H, Zhao J Y, Jia L. Short?time traffic flow prediction. Journal of University of Jinan ( Natural Sciences) 2008, 22(1) : 88~ 94. ( 高慧, 赵建玉, 贾磊. 短时交通流预测方法综述. 济南大学学报(自然科学) , 2008, 22( 1) : 88~ 94) .
[ 2 ] Wang J, Shi Q X. Short?time traffic flow pre diction. Intellisent Transportation System, 2005, 6( 1) : 92~ 98. ( 王?进, 史其信. 短时交通预测模型综述. 智能交通, 2005, 6(1) : 92~98) .
[ 3 ] Yang Y, T ang P, Wang Y C, et al. Neural network structure optimization based on improved LMBP algorithm. Computer Engineering, 2008, 34( 1) : 215~ 217. ( 杨英, 唐?平,
王越超等. 基于 LMBP 改进算法的神经网络结构优化. 计算机工程, 2008, 34(1): 215~ 217).
[ 4 ] Shang N, Qin M G, Wang Y Q, et al. A BP neural network method for short?term traffic flow forecasting on crossroads. Computer Applications and Software, 2006, 23( 2) : 32~ 33.
( 尚宁, 覃明贵, 王亚琴等. 基于 BP 神经网络的路口短时交通流量预测方法. 计算机应用与软件, 2006, 23( 2) : 32~ 33) .
[ 5 ] Kuang A W, Huang Z X. Short?term traffic flow prediction based on RBF neural network. Systems Engineering, 2004, 22(2): 63~ 65. (况爱武, 黄中样. 基于 RBF 神经网络的短时段交通预测模型. 系统工程, 2004, 22( 2) : 63~ 65) .
[ 6 ] Liu X C, Xu Y P, Tang Z L. Improved artificial neural network model for hydrologic time series prediction. Journal of Nanjing University
( Natural Sciences), 2008, 44( 1): 85~ 90. ( 刘星才, 许有鹏, 唐郑亮. 改进的人工神经网络模型在水文序列预测中的应用研究. 南京大学学报(自然科学) , 2008, 44( 1) : 85~ 90) .
[ 7 ] Yang Z S, Wang Y, Guan Q. Short?time traffic flow prediction method based on SVM. Journal of Jilin University ( Engineering and Technology
Edition) , 2006, 36( 6): 881~ 884. ( 杨兆升, 王媛, 管青. 基于支持向量机方法的短时交通流量预测方法. 吉林大学学报( 工学版) , 2006, 36(6): 881~ 884).
[ 8 ] Guo M, Sun Z Q, Pan J S, et al. Research on short time traffic flow forecasting method. Application Research of Computers, 2008, 25( 9):
2676~ 2678. ( 郭牧, 孙占全, 潘景山等. 短时交通流预测方法研究. 计算机应用研究, 2008, 25(9): 2676~ 2678).
[ 9 ] Sun Z Q, Pan J S, Zhang Z J, et al. Traffic flow forecast based on combining principal component analysis with support vector machine.
Journal of Highway and Transportation Research and Development, 2009, 26( 5) : 127~ 131. (孙占全, 潘景山, 张赞军等. 基于主成分分析与支持向量机结合的交通流预测. 公路交通科技, 2009, 26(5): 127~ 131) .
[ 10] She J F, Feng X Z, Du J K. A review on progress in spatiotemporal data modeling. Journal of Nanjing University ( Natural Sciences ),
2005, 41( 3) : 259~ 267. ( 佘江峰, 冯学智, 郁金康. 时空数据模型的研究进展评述. 南京大学学报( 自然科学), 2005, 41(3): 259~ 267).
[ 11] Xie M W. T he relation of covariance, correlation coefficient and correlation. Application of Statistics and M anagement, 2004, 23( 3) : 33~ 36. ( 谢明文. 关于协方差、相关系数与相关性的关系. 数理统计与管理, 2004, 23( 3): 33~ 36) .
[ 12] Guo R Z. Spacial analysis. Bejing: Higher Education Press, 2001. ( 郭仁忠. 空间分析. 北京: 高等教育出版社, 2001).
[ 13] Xu Z X, Lin Y, Li S B, et al. A realtime traffic prediction and pre assessment simulation system dynaCHINA and its parameters calibration.
Shandong Science, 2009, 22(6) : 46~ 49. ( 许兆霞, 林勇, 李树斌等. 实时交通预估预测仿真系统 dynaCHINA 参数标定及应用. 山东科学, 2009, 22(6): 46~ 49).

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed