摘要 随着我国高速公路建设里程的快速增长，特别是西部地区公路网的完善，隧道数量和里程不断增加，隧道路段的交通安全状况不容忽视。隧道特别是长大隧道是公路交通的咽喉，一旦发生事故，对营救和交通组织都极为不利，易引发重大事故和二次事故。因此，隧道路段驾驶行为风险研究受到普遍关注。 本文依托“基于事件的重大公路交通基础设施运营安全研究”课题（国家863高技术研究发展计划资助项目）、“公路隧道及隧道群车辆运行安全保障技术研究”课题（交通部西部交通科技项目）和“增城至从化高速公路隧道（群）安全设计与安全运行控制技术研究”（广州市高速公路有限公司资助项目）。基于隧道路段“人-车-路-环境-管理”动态交通安全系统，以高速公路隧道（群）为研究对象，分析驾驶行为安全影响因素。基于驾驶员在静态运行环境的行车过程中的驾驶需求，提出驾驶员的空间视觉信息需求要素——即空间通视性对驾驶行为的影响。通过研究车-路之间的互动关系，对隧道静态运行环境（路）进行安全评价。分析隧道路段驾驶行为特性，进而提出隧道路段驾驶行为风险评价方法。 首先，对我国高速公路隧道路段交通安全状况进行调研，进而提出安全性等级划分方法，并提出通过对驾驶行为进行研究以改善安全状况。通过分析驾驶行为的安全机理，分析隧道路段驾驶行为的五个影响因素驾驶员、车辆、静态运行环境、环境事件及管理因素，及其相互关系。基于此，为了构建隧道路段驾驶行为研究思路，明确了驾驶行为研究范畴和研究方法。根据隧道驾驶行为研究思路，将驾驶行为风险研究分为四个层次：路网驾驶行为研究，空间通视性驾驶行为研究，风险特征区段驾驶行为研究，风险特征点驾驶行为研究。 其次，驾驶员在静态运行环境的行车过程中的驾驶需求主要表现为视觉需求，也即驾驶员在静态运行环境行车过程中综合体现为对空间视觉信息的需求——即空间通视性。隧道空间通视性是驾驶行为的重要影响因素，即通视与否与驾驶行为的变化规律紧密联系。如果隧道通视，驾驶员认为隧道各环境因素满足驾驶预期，故一般不减速或者减速幅度极小；反之，如果隧道不通视，驾驶员主观认为存在驾驶未知因素，故而驾驶时小心谨慎，在隧道进口路段采取减速行为。通过对隧道空间通视性机理进行分析，提出四个影响因素：纵向搜索范围、几何线形及建筑限界、驾驶员视点及视线偏角位置、人体对明暗过渡环境的视觉特性，并对各因素进行分析。基于以上四个影响因素，提出隧道空间通视性计算模型，得到各因素临界值。结合空间通视性的理论构想和现场实测数据，根据车辆运行车速的变化规律，采用归纳法将隧道分为空间通视型隧道和非空间通视隧道，并给出了各类隧道（群）的风险特征区段划分模型。考虑到隧道群的运行环境更加复杂，所以将隧道群作为一类特殊研究对象。 再次，车-路系统中，隧道路段静态运行环境体现了道路使用者对交通基础设施的需求，是影响驾驶行为的外在环境。通过对隧道驾驶过程中驾驶需求和潜在风险的分析，建立隧道静态运行环境与驾驶行为关系评价指标体系。隧道路段静态运行环境评价对象层包括线形安全性、视距安全性、路面状况安全性、洞内环境安全性、标志标线设施认知性、附属安全设施有效性、监控设施有效性、路网安全设施有效性，并提出各对象层的指标及其评价标准。采用物元分析-层次综合评价法实现对隧道静态运行环境的安全评价。 再次，考虑隧道（群）路段运行环境的特性，提出隧道路段6个风险特征点即驾驶风险突变点，对各风险特征点在不同天气、不同时段下的行车制动距离进行分析，得到各特征点的临界安全车速。在驾驶过程中，驾驶制动的影响因素主要包括：车辆初始速度、驾驶员反应时间、路面附着系数及道路纵坡，对各因素的影响机理进行分析，并提出各因素的取值范围。采用驾驶制动计算模型对各风险特征点的制动临界状态进行分析，得到各工况下的临界安全车速，以及非自由流情况下一定车速下的临界流量，可以通过对隧道路段限速和流量控制达到安全管理的目的。 再次，通过现场试验实测隧道路段的驾驶行为状态，包括各类型隧道（群）不同风险特征区段的加速度分布特性、临界车头时距模型、车道变化率等。并根据隧道路段驾驶行为四层次研究思路和驾驶行为风险特性，提出隧道路段驾驶行为风险评价流程，该流程包括隧道路段驾驶行为风险判别总体结构、驾驶风险计算方法、评价标准及风险判别技术流程，实现对隧道路段的驾驶行为风险的评价和监测。 最后，本文依托增从高速公路凤凰山隧道，将高速公路隧道驾驶行为研究的研究成果运用于工程实践。 总之，通过对我国高速公路隧道安全现状的调查，建立了隧道安全性等级标准。提出了隧道静态运行环境的安全评价指标、评价标准和评价方法，实现对隧道交通基础设施的安全评价。另外，通过研究高速公路隧道（群）路段的驾驶行为特性，建立了隧道驾驶行为风险评价技术，为高速公路隧道的动态运营管理提供了理论依据。 关键词：高速公路隧道，驾驶行为，空间通视性，风险特征区段，静态运行环境，物元分析-层次分析法，风险特征点，凸壳算法，风险评价 ABSTRACT With therapid growth of mileageofthe expressway in China, especially the improvement of theroad network of the westernregion, the number andlength oftunnels are growing, but the traffic safetysituation of the tunnel sections can not be ignored. Tunnel especiallytunnel groups is the key position, if there is an accident, it is hard for rescueand transportorganizations,and easilylead toserious accidentsand secondaryaccidents.Therefore, the drivingbehavior risk of tunnel sectionis ofuniversal concern. This paper funded by projects “Operational safety study of major transportion infrastructures during the incident state” (National Development Project 863 of High-technology Research), “Research on the vehicle operation safety and security technology of expressway tunnel and tunnel group” (the Western Traffic Science and Technology Project), “Security design and safe operation technology of Zengcheng to Conghua expressway tunnel (group)” (Guang Zhou Highway CO.,LTD funded project). Based on dynamic traffic safety system of “driver- vehicle- road- environmental- management”, the expressway tunnel (tunnel group) section is taked as the research object, and the factors affectingdriver behavior are analyzed. And based on the driver demand in the process of drivering in the static operating environment, the space visual information demand of the driver —spatial visual pattern is put forward. Focusing on vehicle-road interaction relationship, the tunnel static operating environment is evaluated. Driver behavior characteristics of the tunnel sections are analyzed, and then the tunnel driving risk assessment method is proposed. Firstly, through the traffic safety survey of the expressway tunnel section, the safety classification methods are proposed, and the driving behavior research is put forward to improve the security situation. By analyzing the formation mechanism of driving behavior, five factors of tunnel driving behavior including driver, vehicle, static operating environment, the environment and management factors and their relationship are analyzed. Based on this, in order to build the research framework of driving behavior of tunnel section, the definition of the driving behavior, research areas and research methods are made. Based on the research framework of driving behavior in the tunnel, the research on driving behavior is divided into four levels: research on road network of driving behavior, research on spatial visual pattern of driving behavior, research on the risk on feature section of driving behavior, the risk on feature points of driving behavior. Secondly, the driver demand in a static operating environment in the driving process is mainly for the visual demand, that the driving process in a static operating environment is comprehensive reflect as the needs of space visual information – the spatial visual pattern. For a spatial visual tunnel, the driver considers if the environmental factors meet his expectation, therefore generally does not slow down or slow down in a minimum rate. For a non- spatial visual tunnel, the driver considers there exists some unknown factors, therefore drive be careful to slow down at the tunnel entrance. Through analyzing the mechanism of space visibility, four influencing factors are put forward: vertical search zone, linear geometry and structure clearance, driver’s view position and deflection angle of sight, the visual characteristics of human body in the transition environment from light to dark. Every factor is analyzed. Based on the four influencing factors above, the spatial visual models in tunnel are proposed, as well as the critical value of the factors. Based on the space visibility theory and the field data, tunnel is divided into two types including the spatial visual tunnel and non- spatial visual tunnel by employing the induction method, and the risk characteristics section dividing models are put forward. Considering the complexity of the static operating environment of tunnel group, so the tunnel group is taken as a special type. Thirdly, in the vehicle-road system, static operating environment of tunnel section reflects the need of road users, which is the external environment affecting the driver behavior. By analyzing driving demand and potential risk in the process of driving in tunnel, the evaluation index system between the tunnel static operating environment and driver behavior is established. Evaluation object layers of static operating environment in tunnel section include: linear geometry security, sight distance security, pavement condition security, inside environment security, cognitive effectiveness of traffic signs and markings, effectiveness of ancillary safety devices, effectiveness of monitoring facilities, effectiveness of road network safety facilities. And index and evaluation criteria of every object layer are put forward. The matter element analysis-AHP method is selected to achieve a comprehensive safety evaluation on the tunnel static operating environment. Fourly, considering the characteristics of static operating environment of the tunnel (group), 6 risk characteristics cross sectionals are proposed, which are the mutations of driving risk. The braking distance of risk characteristics cross sectionals under different weather and different time is analyzed, and the criticality safety speed of risk characteristics cross sectionals is obtained. During the driving process, the driving braking factors are vehicle initial speed, driver reaction time, road adhesion coefficient and road longitudinal slope. The impact mechanism of each element is analyzed, and the domain of each element is put forward. The critical state of each risk characteristics cross sectional is analyzed using the driving brake model. The critical safe speed under different conditions and the critical flow in the case of non-free flow in certain speed are studied, indicating that safety management purpose could achieved by limiting the speed and controlling the flow of tunnel sections. Fively, field tests are used to demonstrate the driver behavior characteristics, including the acceleration distribution characteristics, headway - speed model, and the rate of lane change. According to the four-level structure of driving behavior in tunnel sections and the driver behavior characteristics, the risk assessment method of driver behavior is proposed, including the overall risk discriminant structure of driving behavior , driving risk calculation method, evaluation criteria and processes, which can achieve the risk assessment and monitoring of the tunnel section. In the end, the above theoretical results of driver behavior research are applied in Phoenix Mountain Tunnel of Zengcheng-Conghua expresseay as a case study. In short, based on the Chinese expressway tunnel safety status survey, the tunnel safety level standard is established. The safety evaluation index, evaluation criteria and evaluation method of the tunnel static operating environment are put forward, which can achieve the safety assessment of the transport infrastructure. In addition, by the characteristics study of driver behavior, risk assessment techniques is established, which provides the theoretical basis for expressway tunnel safety management. Key Words: expressway tunnel, driver behavior, spatial visual pattern, risk characteristics section, static operating environment, matter element analysis-AHP, risk characteristics cross sectional, convex hull algorithm, risk assessment. |
博士论文