基于贝叶斯空间相关模型的城市快速路安全影响因素研究

An Analysis on Effective Factors of Safety for Urban Expressways Based on Bayesian Spatial Models

快速路是城市交通的主骨架,其运行安全态势直接影响整体通行效率的发挥。利用线圈检测器数据和由事故视频监控系统采集的2011～2013年共50 069起事故数据,应用贝叶斯空间相关模型,对上海市南北高架、延安路高架、内环和中环4条快速路进行建模分析。为了反映路段间的潜在空间联系,在负二项模型的基础上,加入空间随机影响项建立相邻空间相关模型;进一步改进了该空间随机项的协方差矩阵,建立距离空间相关模型。通过贝叶斯回归方法比较3个模型的回归效果。结果显示,距离空间相关模型具有最优的拟合效果,其方差信息标准（DIC）较负二项模型降低6.04%。由此证明,事故在道路中的分布存在空间关联,且其相关性受距离影响。在道路几何特性影响因素中,路段前后1km内匝道数越多,每公里转角越大,事故风险越大;入口-出口型交织路段事故发生几率较高。日均交通量与事故风险显著正相关。不同快速路对比分析表明内环事故风险偏高,中环事故风险较低。

Expressways are the backbones of the urban traffic network. The safety situation on expressways can directlyaffect the efficiency of traffic. 50 069 accidents occurred in the year 2011 - 2013 is collected by central video monitorsystem in a three-year period. Based on traffic flow data from dual-loop detectors and accidents data, a Bayesian spatialmodel is developed to analyze four expressways in the City of Shanghai; North-South elevated road, Yan’ an elevatedroad. Inner Ring, and Middle Ring. In order to reveal the potential spatial correlation among road segments, an adjacentcorrelationspatial model is developed by adding a spatial random-effect factor to the negative binomial model. Further, aspatial correlation model of distance is developed by improving the proximity matrix of the factor. The Bayesian method isapplied to estimate the parameters and compare the regression results of these three models. The results show that thespatial correlation model of distance is superior in term s of goodness-of-fit. The deviance information criterion （DIC） ofthe spatial correlation model of distance is 6.04 % lower than which in the negative binomial model. It indicates that thedistribution of accidents on expressway segments is spatial correlated and is related to the distance between segments. Ananalysis on road geometric features shows that the risk of accident tends to be higher with a larger number of ramps within1 km before or after the midpoint of segments; the cumulative turning angle per kilometer is larger as well. Weavingsegments between on-ramps and off-ramps have a higher risk of accident. The volume of average daily traffic shows a significantpositive correlation with accidents. Compared with the other two expressways, the risk of accident on Inner Ringis higher; while Middle Ring is relatively lower.