高速公路长大隧道群段驾驶员视觉疲劳致因机理与改善建议

Even though the number of dead and wounded in the long freeway tunnel involved traffic accidents has been decreased over the previous decade, the statistics are still alarming and it has already become a serious social problem in China. Obviously, a great amount of factors contribute to the accidents at these locations and driver's vision fatigue plays an decisive role in causing crashes and facilities. Due to the lonely and boring driving environment in long freeway tunnel group, driver's vision faces significant diffuculty in adapting the unexpected lighting change while entering and exiting the tunnel frequently, which may result in a "visual fatigue" and even lead to accidents. However, the all-around understanding of vision fatigue has still not been realized and thus the effective countermeasures could not be pursued. Therefore, this project analyzes the reported findings and collects the eye movement through experiment so as to judge their variation law with time and identify the excitation and fatigue locations of vision. The definitions of vision fatigue as well as vision fatigue index are determined and then it intends to fully reveal the aggravation process of vision fatigue. Moreover, the effects of vision fatigue on driving performance, pavement illumination and speed are also checked through data mining. Finally, the patterns of lighting stimulus in long tunnel are to be designed to the excitability of vision and the effectiveness of proposed measures are verified through the simulation test. The project team has the necessary driving simulation system, advanced working conditions and rich research experiences in human visual and driving behavior that ensure the smooth progress of this research project till achieving a complete success.

近年来我国高速公路长大隧道群段交通事故数居高不下，且严重性强，已成为严峻社会问题，其中驾驶员视觉疲劳是主要诱发因素之一。高速公路长大隧道群驾驶环境封闭单调，频繁进出隧道致使驾驶员视觉明暗适应困难，产生"驾驶疲劳"，从而增加交通事故潜在风险，但目前我国尚缺乏对整个隧道群段驾驶员视觉疲劳问题的系统理论研究，相关管理政策措施无法跟进。本项目拟充分借鉴已有成果，通过现场测试，采集驾驶员眼动等指标参数，分析其变化规律，甄别其视觉"兴奋点"、"疲劳点"；结合指标测试结果，引入视觉疲劳和视觉疲劳指数的概念，确定度量标准，分析驾驶员视觉疲劳从形成到累积的演化规律，并研究其与驾驶绩效、照度、车速的关系；根据驾驶员视觉"疲劳点"在长大隧道内的分布区间，设置灯光刺激，提高视觉兴奋性，并模拟测试实用效果。项目组拥有驾驶模拟系统等实验平台和工作条件，在视觉特性、驾驶行为等方面具有研究基础，具备完成项目研究的条件。

近年来我国高速公路长大隧道群段交通事故数居高不下，且严重性强，已成为严峻社会问题，其中驾驶员视觉疲劳是主要诱发因素之一。高速公路长大隧道群驾驶环境封闭单调，频繁进出隧道致使驾驶员视觉明暗适应困难，产生“驾驶疲劳”，从而增加交通事故潜在风险，但目前我国尚缺乏对整个隧道群段驾驶员视觉疲劳问题的系统理论研究，相关管理政策措施无法跟进。京昆高速G5陕西部分路段和京沪高速G2等山东部分段，开展了驾驶员视觉特性现场测试，并利用G2港沟枢纽立交至彩石立交区间为原型开发驾驶场景和进行模拟驾驶实验，获取了长大隧道出入口段驾驶员的视点变化(瞳孔直径、视点数、视点持续时间、视觉搜索角度方差)、扫视(扫视数、扫视速度、扫视幅度)、眨眼(眨眼频率、眨眼持续时间、闭眼持续时间)及疲劳感知SSS值等指标。利用现场测试数据，分析了隧道入口接近段、适应段及洞内段驾驶员视觉特性的变化规律，发现其基本经历一个指标值从急剧增大到降低、再到缓慢回升并最终保持稳定的过程，藉此分析了驾驶员明暗自适应规律。依据隧道群环境中2h、3h及4h连续驾驶后驾驶员视觉特性指标的数据变化甄别了其视觉“兴奋点”、“疲劳点”，发现驾驶员个体疲劳感与其瞳孔直径、视点持续时间、眨眼频率、眨眼持续时间及闭眼持续时间正相关，而与视点数、视觉搜索角度方差、扫视数、扫视速度及扫视幅度负相关，进而利用这10项视觉特性指标的变化率构建了疲劳感知SSS变化预测模型。以速度感知、注意力集中及决策反应三项指标界定了驾驶员驾驶绩效的概念，分析了其变化特征，研究结果表明2h、3h及4h连续驾驶后DPI值分别降低了1.89%、5.60%及9.21%，而对驾驶员的速度感知能力影响极小，据此分析了驾驶员视觉疲劳的致因机理。根据实验研究与隧道群事故影响因素辨识的结果，系统研究提出了驾驶员视觉疲劳与安全保障的措施，包括出入口处的减光栅格设置、洞内外照度差限制、平/竖曲线设置、GPS语音导航及连续驾驶时间限制等，相关研究成果可为高速公路长大隧道群段交通安全管理及事故预防提供新思路和新靶点。