基于安全替代指标的城市道路交通安全高风险点位研判方法  

High Risk Spots on Urban Roads Research and Judgment Methods Based on Surrogate Safety Measures

作　　者：姚锐 王俊骅[2] YAO Rui;WANG Junhua

机构地区：[1]上海市政工程设计研究总院(集团)有限公司,上海200092 [2]同济大学,上海200092

出　　处：《上海公路》2023年第3期116-120,125,M0006,M0007,共8页Shanghai Highways

摘　　要：针对城市道路开展高风险点位的判别和筛查,是交通安全改善工作的重要内容。以上海市为例,基于手机导航定位数据,提取了急减速事件、急加速事件、平均速度、速度变异系数和拥堵指数五种替代安全指标,进而分别构建了基于零截尾负二项的事故频率预测模型和基于多元logit的事故严重程度预测模型,可有效识别出城市道路交通高风险点位。模型具备较高的准确率和可解释性。该方法可为精细化的城市道路交通安全治理提供有效的技术支撑。Identifying and screening high risk spots on urban roads plays a vital role in improving road traffic safety.Five surrogate safety measures(rapid acceleration events,rapid deceleration events,average speed,coefficient of variation of speed and congestion index)were se lected based on road geometry data,historical crash data and satellite point positioning data.The accident frequency prediction model based on zero truncation negative binomial and the accident severity prediction model based on multivariate logit were constructed respectively,which can effectively identify high-risk points on urban road traffic.The models have high accuracy and interpretability.In general,the proposed methods are supporting technologies for efficient traffic delicacy management measures.

关 键 词：卫星单点定位数据 城市道路 历史事故安全替代指标 静态事故风险研判 

分 类 号：U492.8[交通运输工程—交通运输规划与管理]