智能网联环境下无信号控制交叉口可穿越间隙分析与预测

Analysis and prediction of traversable gap at unsignalized intersection in intelligent networked environment

无信号交叉口环境下,可穿越间隙会随着车速、流量以及智能网联车辆渗透率的变化而波动。利用驾驶模拟器搭建无信号交叉口环境下的智能网联车辆混行场景,对不同交通流状态下驾驶员的驾驶行为、心理特性进行实验;选取速度、加速度、瞳孔直径以及瞳孔直径变化率作为驾驶员的生/心理指标,结合主成分分析和k-means聚类算法对其进行定性与定量分析;仿真结果表明:在不同交通流状态下,驾驶员的驾驶行为发生了明显的改变,可划分为三个等级:“无感无行为间隙”、“有感无行为间隙”、“有感有行为间隙”。实验结果为智能网联环境下无信号交叉口的控制方法提供了数据及理论基础。

Under the environment of unsignalized intersection,the traversable gap will fluctuate with the change of speed,flow and the penetration rate of intelligent networked vehicles.A driving simulator is used to build the mixed traffic scene of intelligent networked vehicles under the environment of unsignalized intersection,and the driving behavior and psychological characteristics of drivers under different traffic flow conditions are tested.Speed,acceleration,pupil diameter and its change rate are selected as the physiological and psychological indicators of drivers,and qualitative and quantitative analysis is carried out by combining principal component analysis and k-means clustering algorithm.The simulation results show that the driver's driving behavior has changed significantly under different traffic flow conditions.It can be divided into three levels:"senseless without behavior gap","sensory without behavior gap",and"sensory with behavior Gap".The experimental results provide a data and theoretical basis for the control methods of unsignalized intersections in the intelligent networked environment.