基于最优加权组合模型的道路交通事故预测  

Road traffic accidents prediction based on optimal weighted combined model

作　　者：赵玲[1,2] 许宏科[1] 程鸿亮[1] 

机构地区：[1]长安大学电子与控制工程学院,西安710064 [2]西安邮电大学通信与信息工程学院,西安710121

出　　处：《计算机工程与应用》2013年第24期11-15,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.60804049);中央高校基本科研业务费专项资金项目(No.CHD2012JC056)

摘　　要：交通事故预测是交通安全评价、规划和决策的基础。针对各种单一灰色预测模型存在的局限性,建立了一种基于最优加权的灰色组合预测模型。根据我国道路交通事故的发展情况,建立了GM(1,1)、Verhulst和SCGM(1,1)c相结合的组合预测模型,运用最优加权法确定组合预测模型的权重系数。利用2001—2007年我国道路交通事故死亡人数的实际值作为原始数据,构建各个单一预测模型和最优组合预测模型,预测其2008—2010年交通事故死亡人数。预测结果表明,组合预测模型比单一GM(1,1)模型、Verhulst模型和SCGM(1,1)c模型具有更高的预测精度。The prediction of traffic accident is the basis of transportation safety, assessment and decision-making. Aimed at solving the limitations in various single grey forecasting methods, a combined forecasting model of road traffic accidents based on optimal weighted method is put forward. According to the characteristics of traffic accident in China, a model combining GM（ 1,1 ）, Verhulst and SCGM （ 1,1 ） c is established and the weight Coefficients of combined forecasting model are determined by optimal weighted method. The deaths of traffic accident in China from 2001 to 2007 are taken as original data to establish forecasting model predicting the deaths of traffic accident from 2008 to 2010. The results demonstrate that the forecast accuracy of combined model is better than that of GM （ 1,1 ）model, Verhulst model and S CGM （ 1,1 ）c model.

关 键 词：交通安全 交通事故 最优加权 组合预测 

分 类 号：U492.3[交通运输工程—交通运输规划与管理]