高速公路交通流的分形维数与相空间重构预测  

Fractal Dimension and Forecasting of Traffic Flow for Freeway Based on Phase Space Reconstruction

作　　者：李建章[1] 朱顺应[1] 

机构地区：[1]重庆交通大学,重庆400074

出　　处：《重庆交通大学学报（自然科学版）》2007年第6期119-122,共4页Journal of Chongqing Jiaotong University(Natural Science)

基　　金：交通部应用基础研究项目(200331981408)

摘　　要：对成渝高速公路短时交通流通过计算不同时间尺度下Hurst指数而等到其相应的分形维数,结果表明,时间间隔越短的交通流,其分形维数越大,结构越复杂.由于时间间隔越短的交通流随机性大和复杂的结构,所以预测也就越困难.提出了一种新的基于相空间重构和移动平均相结合的预测方法——移动平均最近邻域法,从理论与实际数据两方面分析和验证了该方法对短时交通流预测的有效性.By calculating the Hurst Index of the short time traffic flow in variant time scales for freeway, a fractal dimension was put forward. It showed that the interval time of traffic flow was shorter, the fractal dimension was bigger, and the structure was more complex. So it is more difficult to forecast. We present a new forecasting method - moving average nearest neighborhood forecasting method, which is based on the phase space reconstruction and combining with the moving average forecasting and the nearest neighborhood method. The efficiency of the method was proved by theory and actual traffic flow.

关 键 词：高速公路 交通流预测 HURST指数 分形维数 相空间重构 移动平均最近邻域法 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]