城市快速路交通流分形预测算法的研究  

Research of Fractal Forecasting Algorithm of Traffic Flow on Urban Expressway

作　　者：梅宏标[1,2] 王坚[1] 张慧哲[1] 

机构地区：[1]同济大学CIMS研究中心,上海201804 [2]江西理工大学,江西赣州341000

出　　处：《公路交通科技》2009年第10期105-110,共6页Journal of Highway and Transportation Research and Development

基　　金：国家科技支撑计划基金资助项目(2006BAF01A46);上海市社会发展重大专项基金资助项目(06DZ12001);上海市科技发展基金重大项目(06DZ12001);上海市科技发展基金重点项目(061612058)

摘　　要：根据城市快速路交通网络几何结构及交通流的混沌性,在分析交通流分形特性的基础上,应用小波系数方差法求解Hurst系数,同时建立了城市快速路交通流分形预测模型。利用实测数据,运用分形预测模型对上海市高架路的交通流进行了预测及误差分析,探讨了交流量不同状态下,即不同时段相同时间尺度,同一时段不同时间尺度;同一时段同一时间尺度下不等长数据序列的交通流分形特性及预测结果,并与传统方法进行比较,验证了模型的有效性及正确性。Based on the research of the geometric structure of traffic network, chaos and fractal characters of traffic flow of urban expressway, the Hurst coefficient was solved by using wavelet coefficient variance algorithm, and a fractal forecasting model was established. Traffic flow forecasting and error analyzing of the viaducts in Shanghai were conducted based on the fractal forecasting model. The characters of traffic flow fractal and the results of forecasting under different states （different time-span but same time-scale, same time-span but different time-scales, same time-span and same time-scale but different data sequences） were discussed and compared with traditional algorithms, and the validity of this model was proved.

关 键 词：交通工程 分形 小波 城市快速路 自相似性 交通流 

分 类 号：U491[交通运输工程—交通运输规划与管理]