基于ARIMAX模型的交通事故宏观预测  

Traffic accident macroscopically forecast based On ARIMAX Model

作　　者：李春燕[1] 陈峻[1] 

机构地区：[1]东南大学交通学院,南京210096

出　　处：《道路交通与安全》2009年第1期18-22,共5页Road Traffic & Safety

基　　金：教育部新世纪优秀人才支持计划(NCET-07-0176)

摘　　要：针对现有道路交通事故预测方法的不足,采用逐步回归法从众多宏观影响因素当中筛选出主要影响因素,并将动态回归ARIMAX模型应用于预测。前者保证了模型应用的准确性,后者则兼有回归与时间序列预测方法两方面的优点。根据1983年—2005年间相关数据,建立起道路交通死亡人数同人口总数、运输线质量里程数、客运量、驾驶员人数、人均GDP、公路运输汽车拥有量的相关关系,进一步应用ARIMAX模型进行预测,拟合结果显示,误差较小、预测情况良好,在交通事故宏观预测方面有很好的应用前景。Aimed at the deficiency of traffic accident estimation methods in existence, it firstly selected the main impact factors during many macroscopical ones used stepwise regression method and applied ARIMAX mode which is a dynamic regression method to forecast. The former ensures the veracity of method application, while the latter owns both regression and ETS＇s merits. According to the related data during 1983 and 2005, it founded the relationship between the dead people due to traffic accident and total population, mileage of road rank, passenger transport number, the population of drivers, the average GDP and automobile numbers. And it forecasted with ARIMAX method future. The result shows that the error is small and the forecast data is better and has a much better foreground on the aspect of traffic accident macroscopically forecast.

关 键 词：逐步回归 ARIMAX模型 筛选因素 宏观预测 

分 类 号：U491.31[交通运输工程—交通运输规划与管理] TV214[交通运输工程—道路与铁道工程]