摘要
为解析车型差异对主线汇入间隙选择特征的影响,同时考虑到传统的单一模型拟合分布方法存在模型选择不确定性问题,应用贝叶斯模型平均(BMA)方法进行分车型汇入间隙分布建模:采集高速主线合流区车辆轨迹数据,提取四类(依车型划分)交互车辆对时间与空间滞后间隙样本,构建候选分布概率密度函数模型空间;采用贝叶斯信息准则近似计算模型后验概率,综合加权得到BMA分布模型。结果表明:①主线汇入滞后间隙分布建模过程存在显著模型选择不确定问题,BMA方法可有效融合多模型优势,提升拟合稳健性;②滞后车类型与滞后间隙分布具有强相关性;③滞后车为小型车时,大型车汇入时较小型车汇入更有可能选择更小空间滞后间隙。
To investigate how vehicle type heterogeneity influences lag gap selection behavior during freeway mainline merging––and to overcome the model selection uncertainty inherent in conventional single-model distribution fitting––this study adopts a Bayesian Model Averaging(BMA)framework to characterize lag gap distributions under different vehicle-type interaction conditions.Using highresolution vehicle trajectory data collected in a freeway merging area,time and space lag gap samples were extracted for four categories of interacting vehicle pairs,classified according to vehicle type.A candidate distribution model space consisting of multiple probability density functions was constructed,and the posterior model probabilities were approximated using the Bayesian Information Criterion(BIC).The final BMA-based distributions were obtained by weighting and combining the candidate models according to their posterior probabilities.The findings reveal that:①substantial model selection uncertainty exists in the distributional modeling of lag gaps,and the BMA effectively integrates the advantages of multiple models,thereby improving robustness and fitting accuracy;②the type of lag vehicle exhibits strong correlation with the statistical characteristics of lag gaps;③when the lag vehicle is a small vehicle,large vehicles tend to merge with smaller space lag gaps than small vehicles under comparable conditions.
作者
廖彦锟
邹亚杰
吴淑博
李林波
LIAO Yankun;ZOU Yajie;WU Shubo;LI Linbo(College of Transportation,Tongji University,Shanghai 201804,China;The Key Laboratory of Road and Traffic Engineering,Ministry of Education,Tongji University,Shanghai 201804,China)
出处
《交通与运输》
2026年第1期25-32,共8页
Traffic & Transportation
基金
国家自然科学基金(52472351)
上海市国际科技合作基金项目(24170742100)
中央高校基本科研业务费专项资金(22120230311)。
