无信号交叉口横向冲突预警对驾驶行为的影响

Impact of early-warning for lateral conflict at unsignalized intersections on driving behavior

针对智能网联汽车在无信号交叉口对横向冲突的避撞问题,本研究设计了横向冲突人机交互预警及横向冲突驾驶模拟试验场景,并分析不同类型预警对驾驶行为的影响规律。运用驾驶模拟器搭建城市道路无信号交叉口并设计横向冲突事件,选取30位驾驶人作为试验对象,应用眼动仪、驾驶模拟平台开展驾驶模拟试验;采集风险场景下驾驶人的眼动、操作行为与车辆运行数据;对比分析各组场景下驾驶人的注视时长和扫视频率及行车速度、制动时间和踩制动踏板速率等驾驶行为特性参数,分别运用方差分析和描述性统计方法分析了预警提示类型和提示时机两个方面对驾驶行为特性的影响。结果表明:网联警告型预警的警示效果优于常规提示型预警,但该类预警使驾驶人操作车辆时的减速度增大至-4 m/s^(2)以上,导致行车舒适性降低;预警提示时机对驾驶行为的影响显著,且预警提示位置宜设置在距离交叉口50~90 m处。研究结果能够为交通安全设施开发者提供有效HMI预警设计建议,提高无信号交叉口通行安全性。

This study addresses the problem experienced by intelligent connected vehicles of avoiding lateral conflict collisions at unsignalized intersections.It involves the design of a human-machine in-teractive early-warning system for lateral conflicts,and the creation of a lateral conflict driving-simu-lation test scenario to analyze the impact of different types of early warnings on driving behavior.A driving simulator was utilized to construct an unsignalized intersection on an urban road in which a lateral conflict event was simulated.Thirty drivers were selected as test subjects.An eye tracker and driving-simulation platform were employed to facilitate the experiment.Data on drivers’eye move-ments,operational behavior,and vehicle operation were collected under different risk scenarios.Pa-rameters such as driver’s fixation duration,saccade frequency,driving speed,braking time,and ped-al speed were compared and analyzed.Descriptive statistics and variance analysis methods were uti-lized to analyze the effects of early warning types and timing on driving behavior characteristics.The results show that early warning systems with Internet connectivity are more effective in producing warnings than traditional reminder-based early warning systems.However the former increases the deceleration rate during vehicle operation to more than-4m/s^(2),reducing driver comfort.The timing of early-warning prompts significantly affect driving behavior,the prompts should be set at a dis-tance of 50 to 90 m before the intersection.The study results offer valuable suggestions for traffic safety facility developers for the design of Human-Machine Interface(HMI)warnings to enhance safety at unsignalized intersections.