信号交叉口基于出发时刻预测的生态驾驶方法

在城市道路交通中,信号交叉口区域内车辆频繁停车启动的现象,加剧了整体交通流的能源消耗、污染排放与车辆延误。为了减少信号交叉口启停波现象对整体交通流产生的负面影响,以面向未来人工驾驶车辆(HDV)/智能网联车辆(CAV)混合构成的新型混合交通环境为基础,提出了一种基于出发时刻预测的生态驾驶方法,通过优化CAV的驾驶轨迹,减少交叉口区域的车辆延误和能源消耗。首先,对混合交通流的基本图模型进行了分析,根据启停波影响范围,划分CAV通过交叉口的驾驶场景;然后,建立了子区渗透率对饱和车头时距的影响关系,预测了CAV以当前饱和车头时距通过交叉口的时间;最后,结合车辆与交叉口的距离,利用分段三角函数模型,生成其通过交叉口的速度限制曲线,并将优化速度嵌入到智能车辆的跟驰模型中作为限制速度,从而使CAV在无法通过当前绿灯窗口的条件下,实现提前减速,在通过交叉口区域后解除速度限制,切换回自身的跟驰模型。此外,还提出了平均综合效能这一指标来综合评价驾驶策略在效率和能耗2个方面的性能,并将提出的基于出发时刻预测的生态驾驶方法与传统网联车辆控制方法、经典交叉口节能控制方法进行了对比。研究结果表明:提出的出发时刻预测方法可以精确预测CAV在交叉口的出发时刻,有效减少车辆的能源消耗与污染排放,同时提高信号交叉口的通行效率;在渗透率大于60%情况下,该方法对系统效能的提高达到12%左右,在10%渗透率条件下也可以达到6%的效能增益;在交通饱和流率在0.5~0.9的范围内时,系统的效能增益较明显。

Eco-driving Method for Signalized Intersection Based on Departure Time Prediction

In urban traffic systems, frequent parking and activation of vehicles in signalized intersections exacerbates energy consumption, pollution emissions, and vehicle delays in the overall traffic flow. To reduce the negative impact of the start-stop wave at signalized intersections on the overall traffic flow, a new mixed traffic environment comprising a hybrid of human driven vehicles and connected and automated vehicle (CAVs)-a Departure Time Prediction (DTP) based eco-driving method-is proposed. By optimizing the driving trajectory of the CAV, the vehicle delays and energy consumption are reduced in the intersection areas. First, the mixed traffic flow fundamental graph model was analyzed, and the driving scenarios of CAVs passing through the intersection were calculated according to the impact range of the start-stop wave. Then, the relationship between the penetration rate of sub-region and saturated time headway of the vehicle was established; subsequently, the time when the CAV passes through the intersection was predicted. Finally, by combining the distance between the vehicle and the intersection, a piecewise trigonometric function model was used to generate the vehicle's speed limit curve; the optimized speed was embedded into the car following model as the limit speed. As a result, CAV achieved early deceleration under the condition that it cannot pass the current green light window, removed the speed limit after passing the intersection area, and switched back to its own car-following model. In addition, an evaluation index-average comprehensive efficiency-is proposed to comprehensively evaluate the performance of the driving strategy in terms of efficiency and energy consumption. The proposed DTP method was compared with the traditional CAV control method and classic intersection eco-driving method. The research results demonstrated that the proposed starting time prediction method could accurately predict the passing time of a vehicle in the next green light signal window. CAVs can slow down in advance after entering the communication range of the intersection while avoiding the impact of deceleration waves in the intersection area. This method can effectively reduce the energy consumption and pollution emissions of the vehicles. Simultaneously, it can improve the traffic efficiency of signalized intersections by weakening the impact of deceleration waves on mixed traffic flows. When the traffic penetration rate is greater than 60%, the method improves the system efficiency by approximately 12%; additionally, an efficiency gain of 6% can be achieved when the penetration rate is 10%. The performance gain of the system is more evident when the traffic saturation flow rate is in the range of 0.5-0.9.