摘要
针对面向全车型的道路拥堵状态评估统一化和静态管控方法的局限性,围绕不同类型车辆的交通拥堵问题开展研究。首先,基于沈海高速公路福建段龙门架数据,采用聚类方法分析不同类型车辆在不同指标下的交通拥堵特征;其次,提出分车型数据驱动的模糊云模型,根据行业专家给出的数字特征计算不同车型下的拥堵等级隶属度;然后,选取拥堵指标作为模糊逻辑系统的输入,并采用Mamdani方法建立模糊关联矩阵,量化不同专家评价的可信度;引入Dempster-Shafter证据理论对所有专家的综合可信度进行模糊融合,根据隶属度最大原则综合评判道路交通状态;最后,验证了提出模型的精度及对车型间差异表征的准确性,分车型下的拥堵评价结果一致性达84.51%。结果表明:相比1~4型客车及小型货车,大型货车在移动过程中形成的瓶颈对高速公路上堵塞的消散表现出明显的迟滞作用;在低密度下,客车、少轴货车、多轴货车渗透率的提高对路况的影响依次提升;在高密度下,3~6轴货车渗透率的改变对拥堵状态的影响最为明显,1~4轴客车次之。分车型拥堵研究可为制订适应各类车型的主动交通管理提供定量的数据支撑,特别是在动态调控措施如错峰出行、可变限速等方面具有重要指导意义。
For the limitation that the unified static road congestion state assessment and control method for all vehicles are not refined,this paper focuses on the traffic flow of different types of vehicles.First,based on the gantry data of Fujian section of Shenyang-Haikou expressway,the traffic congestion characteristics of different types of vehicles with different indicators were analyzed by using clustering method.Second,the data-driven fuzzy cloud model based on different vehicle types was established to calculate the affiliation degree of congestion levels according to the numerical characteristics given by experts.Then,the congestion indicators were selected as the input of fuzzy logic system,and the Mamdani method was used to establish the fuzzy correlation matrix to quantify the credibility of different experts’evaluation.Subsequently,the Dempster-Shafter evidence theory was introduced to fuzzy fuse the comprehensive credibility of all experts.The road traffic state was judged comprehensively according to the maximum affiliation principle.Finally,the model precision and the accuracy of representation on differences among vehicle types were verified.The consistency of congestion evaluation results based on different vehicle types reached 84.51%.The result indicates that compared with 1-4 types of passenger cars and mini trucks,the bottlenecks formed by large trucks during movement exhibit a significant hysteresis effect on the expressway congestion dissipation.Under low density,the influence of penetration rate increasing of passenger cars,less-axle trucks and multi-axle trucks on road conditions is sequentially enhanced.Under high density,the penetration rate variations of 3-6 axle trucks have the most obvious influence on congestion state,followed by 1-4 axle passenger cars.The study on congestion based on different vehicle types can provide quantitative data support for active traffic management of various types of vehicles,especially has important guiding significance in dynamic regulation measures,e.g.,sta
作者
李彬
牛万龙
胡纪龄
文江辉
LI Bin;NIU Wan-long;HU Ji-ling;WEN Jiang-hui(School of Automobile,Chang’an University,Xi’an,Shaanxi 710018,China;School of Science,Wuhan University of Technology,Wuhan,Hubei 430070,China)
出处
《公路交通科技》
CAS
CSCD
北大核心
2024年第8期40-54,共15页
Journal of Highway and Transportation Research and Development
基金
国家自然科学基金项目(52272354)
国家重点研发计划项目(2022YFE0102700)。
关键词
智能交通
高速公路
拥堵判别
模糊云模型
分车型
intelligent transport
expressway
congestion discrimination
fuzzy cloud model
different vehicle types
