单车场景下城市交叉口的智能驾驶车辆左转决策研究

在复杂动态的城市道路环境中,不同的交通参与者之间会不可避免地产生时间或空间上的冲突。针对该问题,对智能驾驶车辆在城市交叉口左转时潜在的冲突行为进行分析并建立决策模型。考虑了车辆运动模式并基于高斯过程回归模型(GPR)建立了直行车辆长时轨迹预测模型,结合轨迹预测提出了基于冲突消解的智能驾驶车辆决策流程(模型)和考虑多因素的驾驶动作选择方法。基于Matlab/Simulink&Prescan搭建仿真验证平台,联合真实数据对算法进行验证。结果表明,单车场景下,决策模型能够以90%的成功率引导无人驾驶车辆完成通行任务。

Research on Decision-Making Model for Left-Turning Autonomous Vehicles at Urban Intersections

The left-turn decision-making model with the consideration of surrounding vehicles’information in complex and dynamic urban environments is critical for the development of autonomous vehicles.In this paper,we propose an approach to improve levels of safety for autonomous vehicles to cross urban intersections.Firstly,the trajectory data were collected by the subgrade strain sensors and a retrofit autonomous vehicle.Then,the GPR model was trained to predict the trajectories of oncoming vehicles going straight.In addition,a conflict resolution method was proposed and a decision maker for selecting driving maneuvers was formulated considering multiple factors,such as safety,efficiency,comfort and altruism.Finally,a simulation and verification platform was built using Prescan and MATLAB Simulink.The reliability and effectiveness of the tactical-level decision-making model was verified by both simulation results and real-road experimental results.The paper provides a theoretical basis for the in-depth study of left-turn driving decision-making in complex and uncertain environment.