信号交叉口行车风险场建立及车辆通行行为优化

随着汽车逐步向智能化、网联化发展，智能网联车辆逐步进入实际应用阶段。进行智能网联车辆的通行行为优化，对提升驾驶安全性和行车效率，避免事故发生和交通拥堵至关重要。车辆在通过交叉口时将受到很多环境及运动因素的影响，而现有的通行优化模型难以准确表达各类因素共同作用下的行驶环境。为此，基于风险场理论建立由环境场和运动场组成的信号交叉口行车风险场，表征信号交叉口中每点的实时行车风险程度，从而引导车辆驶向风险值低点，并提供下一步长的位移及速度指引，实现车辆的动态轨迹优化及速度控制。典型场景下的仿真结果表明：在优化模型的控制下单车的信号交叉口通行效率明显提升，其中直行方向车辆单车平均通行效率提升最高，平均提升6.35%,通过对交叉口面积内所有车辆进行通行行为优化，交叉口通行效率提升了9.3%,这表明所建模型可以准确表达交叉口行车环境并优化车辆通行行为。研究结论可应用于自动驾驶车辆的交叉口通行控制，并为网联环境下的行车环境表达和安全驾驶控制提供模型基础。

Construction of Risk Field and Optimization of Driving Behaviors for Signalized Intersections

With the intelligent and connected development of vehicles, Intelligent Connected Vehicles (ICVs) have gradually entered the practical application stage. It is of great importance that optimizing the driving behaviors of ICVs, which will improves the driving safety and efficiency, and benefits the prevention of traffic accidents and congestion as a result. There are many factors that have impact on vehicles driving through an intersection. They can be basically divided into two types, i.e., environment factors and moving objects. However, the previous studies fail to express these factors accurately. This paper constructs a driving risk field for signalized intersections, which includes environment field and moving vehicle field.Thereal-time driving risk degree for each point in an intersection is characterized as a value in the field. Then the position and speed guidance is provided to the vehicle, which navigates it to the lowest risk point for the next step. The simulation results indicate that the average efficiency of each simulation vehicle is significantly improved, in which, that of the straight direction vehicles increased the most, which is 6.35%. The overall traffic efficiency of all the simulation vehicles increased by 9.3%.This reveals that the proposed method gives a proper expression of the driving environment and improves the driving behaviors for vehicles passing through signalized intersections. This work is devoted to the driving control of autonomous vehicles under signalized intersection scenario. It also provides a potential contribution to the driving environment expression and safety control in network environment.