能效均衡的网联交叉口车辆纵向控制策略研究

The intersection, as crucial points for interrupted and continuum traffic flow transforming into each other, always lead to the increase of energy consumption and the decrease of vehicles operating efficiency due to the stop-and-go driving and acceleration and deceleration behavior. Because of the complexity of the vehicle behavior, the design of the traffic control system is difficult and the stability is not sufficient. This project investigates the topic of "how to improve the traffic capacity at the intersection and minimize the energy consumption of the cooperative vehicles”. This includes: 1) to analyze the characterization of multi-vehicle space-time information, signal timing factor, and vehicle automatic control parameters on car-following behavior, a multi-vehicle information fusion car-following model under the influence of signal is proposed based on the minimum safety distance model; 2) a dynamic traffic flow model is established based on the proposed car-following model, and the disturbance injection method is used to derive the maximum traffic flow cooperative control guidelines; 3) based on the emission model and traffic flow control rules, combining with boundary constraints and path constraints, we propose a dual target vehicle Longitudinal control model with short parking time and low fuel consumption, and then design a low complexity optimal solution. The research work of this project is helpful to build an efficient energy saving optimization control system for cooperative vehicle at the intersection, so it has both theoretical interest and practical value.

交叉口作为城市路网中连续交通流与间断交通流相互转化的关键节点，频繁的车辆启停及加减速变换行为极易引发能源消耗及通行率降低。由于交叉口车辆行为特性复杂，交通控制系统的设计难度较大且稳定性不高。本课题围绕“如何在车联网环境下降低交叉口车辆燃油消耗同时提高通行效率？”展开研究，具体内容包括：1）引入多车时空信息、信号灯配时因素与车辆自动控制响应参数，对现有最小安全距离模型进行分析并完善，建立信号灯影响下多车感知信息融合的车辆跟驰模型；2）根据跟驰模型构建动力学车流模型，并引入扰动注入法，推导通行最大化车流协同控制准则；3）基于交通流控制规则，结合车辆牵引力与驾驶舒适度约束，以燃油消耗为代价函数，构建停车时间短与燃油消耗低的车辆纵向控制模型，并设计低复杂度最优求解方法。本课题的研究工作有助于构建高效的车-路协同交叉口节能优化控制系统，因此具有重要的理论价值和实际意义。

交叉口作为城市路网中连续交通流与间断交通流相互转化的关键节点，频繁的车辆启停及加减速变换行为极易引发能源消耗及通行率降低。由于交叉口车辆行为特性复杂，交通控制系统的设计难度较大且稳定性不高。本课题围绕“如何在车联网环境下降低交叉口车辆燃油消耗同时提高通行效率？”展开研究，具体内容包括：（1）引入多车时空信息、信号灯配时因素与车辆自动控制响应参数，分析并优化标定现有多种经典跟驰模型和基于数据驱动的跟驰模型，建立信号灯影响下多车感知信息融合的车辆跟驰模型，实验以速度、轨迹、磁滞曲线、箱线图等指标，对比分析了7种跟驰模型在信号交叉口处的停走/滑行两种驾驶行为的跟驰效果；（2）根据跟驰模型构建动力学交通流模型，引入扰动注入法，分析交叉口交通震荡机理，理论推演交通流稳定性指数判别表达式，实验分析了不同网联车辆渗透率下车辆队列的局部稳定性和渐进稳定性，推导通行最大化交叉口协同控制准则。（3）在整个交叉口速度引导区，建立了考虑燃油消耗与通行时间的多目标优化车辆纵向控制模型，设计低复杂度的精英遗传算法的多目标优化问题求解方法，仿真验证了车辆在绿灯和红灯相位下的通行时间、燃油效果以及算法求解性能。进一步的，考虑了驾驶员对瞬时速度的响应差异，提出了基于PID控制的网联交叉口车辆队列速度优化方法，通过三角函数规则法优化控制队列头车速度，并利用跟驰模型引导车队其他车辆，实验仿真了不同智能网联汽车市场渗透率下的车辆队列能耗排放与通行效率，验证了策略的有效性。本课题的研究工作有助于构建高效的车-路协同交叉口节能优化控制系统，具有重要的理论价值和实际意义。