基于混沌理论的交通流短时预测模型  

A Chaos Theory-Based Model of Short-Time Forecasting of Traffic Flow

作　　者：李洪萍[1] 裴玉龙[1] 

机构地区：[1]哈尔滨工业大学交通科学与工程学院,黑龙江哈尔滨150090

出　　处：《昆明理工大学学报（理工版）》2006年第5期95-99,105,共6页Journal of Kunming University of Science and Technology(Natural Science Edition)

基　　金：教育部博士点基金项目(项目编号:20030213030);哈尔滨工业大学校基金资助项目(项目编号:HIT.2002.76)

摘　　要：交通流预测是交通系统可行性分析、交通设计和交通管控的基础,短时预测是交通流预测的难点.论文在分析现有交通流预测方法的基础上,提出了一种基于混沌理论的交通流短时预测方法,利用基于小数据量的W olf改进算法计算了流率序列的最大Lyapunov指数.将基于Lya-punov指数的一维预测模式具体化,建立了交通流短时预测模型,并对模型进行了改进,改进后的预测结果具有较高的精度.该模型在智能交通系统(ITS)的交通控制与诱导方面具有广阔的应用前景.Traffic flow forecasting is the basis of feasibility analysis of traffic system, traffic design, traffic management and control, in which the short - time forecasting is more difficult. The existing traffic forecasting models are analyzed, and a new short - time forecasting model based on chaos theory is developed. The Largest Lyapunov Exponent of flow rate series is estimated with the improved Wolf arithmetic method based on small data sets. A short - time forecasting model is developed from the one - dimensional forecasting mode based on the Largest Lyapunov Exponent. The forecasting precision of the improved model is relatively high and the model has a wide potential in traffic control and inducement in ITS field.

关 键 词：交通流时间序列 混沌 LYAPUNOV指数 短时预测模型 

分 类 号：U491[交通运输工程—交通运输规划与管理]