人机混驾交通流事故预警换道决策及特性分析

Lane-changing Decision and Characteristic Analysis on Accident Warning of Mixed Traffic Flow with Human-driving and Autonomous Vehicle

为分析混有网联自动驾驶车辆(CAV)和人工驾驶车辆的混合交通流在事故车辆影响下的交通流特征,针对高速公路发生事故后导致车道临时关闭的场景,研究了人机混驾交通流换道决策机制以及车联网环境下事故预警长度对道路通行能力的影响。基于Gipps跟驰模型和加州大学伯克利分校提出的自动驾驶模型构建了双车道混合交通流元胞自动机模型。分别构建了事故预警区人工驾驶车辆和CAV的倾向型和强制型换道规则,引入了渗透率和事故预警区长度两个参数,分析了混合交通流在事故瓶颈区的运行特征。利用元胞自动机模型对其进行数值模拟,研究其对瓶颈区交通状态演化规律的影响。结果表明:当事故发生时,瓶颈区通行能力随自动驾驶车辆渗透率的增大而提升,当渗透率达到40%~60%时,通行能力约提升13%~38%;当流量水平处于1080~2880 veh/(h·ln)时,事故预警换道机制能有效提高事故发生路段瓶颈区的通行能力,相同渗透率下车辆的平均延误随预警区长度的增加而降低并趋于稳定,当渗透率为40%时,事故最佳预警长度为550~1300 m,且随着渗透率的提高,延误水平降低效果更显著。本研究可为未来人机混驾环境下事故预警策略以及事故处理机制提供理论依据。

In order to analyze the characteristics of the mixed traffic flow with a mixed network of connected and autonomous vehicles(CAV)and human-driving vehicle under the influence of an accident vehicle,the scene of temporary lane closure after an accident on an expressway is analyzed,and the lane-changing decision-making mechanism of human-vehicle mixed traffic flow and the effect of accident warning length on road capacity are studied.Based on the Gipps car-following model and the autonomous driving model proposed by the University of California,Berkeley,a two-lane hybrid traffic flow cellular automata model is constructed.The predilection and mandatory lane-changing rules of manual vehicle and CAV in accident warning area are constructed,two parameters of permeability and accident warning area length are introduced,and the operation characteristics of mixed traffic flow in accident bottleneck region are analyzed.The cellular automata model is used to simulate the mixed traffic flow and the influence of traffic flow on traffic evolution in bottleneck area is studied.The result shows that(1)the traffic capacity of the bottleneck area increases with the increase of the penetration rate of autonomous vehicles,and increases by 13%-38%when the penetration rate reaches 40%-60%;(2)when the traffic flow level is between 1080 and 2880 veh/h/ln,the traffic capacity of the bottleneck area can be effectively improved by the accident warning and lane-changing mechanism,under the same permeability,the average delay of vehicles decreases and tends to be stable with the increase of the length of the warning zone.When the permeability is 40%,the optimal warning length is 550-1300 m,and with the increase of the permeability,the average delay of vehicles will decrease,the effect of delay level reduction is more significant.The study can provide a theoretical basis for accident warning strategy and accident handling mechanism in the future human-vehicle mixed driving environment.