摘要 随着我国高速公路建设里程的快速增长，逐步完善的高速公路网加快了不同地域间物质和人员的流动，在经济与社会的发展中发挥着举足轻重的作用。由于我国高速公路的管理手段和管理体制滞后于建设里程的增长速度，导致高速公路在实际运营过程中存在一系列的安全问题。同时，近年来我国多次遭受冰雪和暴雨等恶劣天气，导致大范围的交通堵塞或中断等事件。因此，实际的高速公路安全管理需求、以及计算机信息化而引发的技术革命，迫切需要对事件状态下高速公路的运营安全进行系统地、科学地研究，以便建立科学而有效的安全管理体系，从而加快高速公路交通安全管理技术的革新过程。 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）和“恶劣气象条件下山区公路网应急事件快速评估与交通组织技术”课题（“十一五”国家科技支撑计划），以高速公路为研究对象，以典型的恶劣天气和应急事件为前提条件，基于运营风险分析与管理的理论，研究高速公路路段级（点/线）和路网级（面）两个层次的运营安全，系统地解决事件状态下微观交通流和中观交通流的运营安全评价问题。 首先，对雨、雾、冰和雪等灾害天气的类型和特征进行分析，构建不同天气条件下恶劣气象耦合事件组合。依据浓雾对驾驶员行车视距的影响，提出雾天驾驶延迟反应模型，量化不同能见度范围对驾驶员的安全反应时间影响。基于传统的车辆行驶理论，数值模拟分析不同恶劣天气条件下车辆的抗侧滑和抗侧翻性。同时，考虑交通事故、交通事件及自然地质灾害等应急事件对道路产生的事件后果特征，确定高速公路典型事件的风险等级大小。 其次，分析自由行驶、跟车行驶和变换车道行驶等三种车辆基本行驶状态，依据恶劣天气对路面和能见度产生的影响特征，分析不同状态下车辆的安全约束条件和控制标准。其中，基于车辆变换车道的轨迹特征，提出车辆变换车道的几何描述模型，从而整合了跟车安全模型和变换车道安全模型。 再次，依据动态系统的特征及风险分析的方法，提出连续交通流和阻塞交通流两种交通状态下风险分析的技术思路，前者以严重事件后果下小概率事件的后果预防为主，后者则是以轻微事件后果下大概率事件的频率控制为主。考虑不同行驶状态下可能对人体造成的伤害程度，构建不同运营状态下交通流的风险等级。在此基础上，提出服务于路段车辆行驶状态检测的检测器布设技术标准。 再次，依据实际交通流的非均质性和各向异性的重要特征，提出判别交通流是否相关的动态归类技术，制定了直接关联归类和间接关联归类的车辆影响判别法则。通过构建交通流的关联矩阵和风险度矩阵，建立连续交通流的运营风险指数，用以评价事件状态下连续交通流的运营安全性。同时，依据事发路段交通流状态和实际的交通供需关系，选取平均车头时距差和剩余累积流量两个指标，评价事件状态下阻塞交通流的运营风险。 最后，依据公路网运营安全的管理层次和目标，在路段微观运营风险指标的基础上，提出路段动态风险饱和度的中观评价指标，用以评价路网级的运营安全。考虑事件状态下公路网环境的时变特性，以及道路使用用户和管理用户的技术需求，提出事件状态下路网运营安全均衡模型，在最快疏散公路网交通流的同时，确保交通重分配后的路段能满足运营安全的容量约束，并以此形成路段和节点的限速和限流等交通安全管理措施。依托京珠高速河南段的高速路网，实例分析正常状态和事件状态下公路网的交通安全组织技术。 总之，从路段和路网两个层次，研究了事件状态下高速公路（网）交通流的运营风险判别技术，初步建立了事件状态下运营风险的识别与判别技术，为实际高速公路（网）的运营安全管理提供了一定的理论依据。 关键词：高速公路，灾害事件，连续交通流，阻塞交通流，运营安全，风险评价，动态归类方法，变换车道模型，动态风险饱和度，运营安全均衡 ABSTRACT With the rapid growth of expressway construction in China, the developing expressways network accelerated gradually mobile speed of logistics and people between different regions, which played a significant role in the development of economy and society. As the development speed of highway management technology lagged behind the growth speed of expressway construction, there existed a series of safety problems in realistic expressway operation. At the same time, China suffered from all kinds of disastrous weather, such as heavy snow and rains, in these years, which leaded to large-scale disasters such as traffic congestions or interruptions. Therefore, realistic technology needs for expressway safety management and were put forward because of electronics revolution caused by computer information. By founding scientific expressway safety system during different kinds of incident state, a scientific and effective system of safety management system would be established, which can speed up technology innovation of expressway safety administration. This paper founded by one project “Operational safety study of major transportation infrastructures during the incident state” ( National Development Project 863 of High-technology Research) and the other project “Rapid assessment and traffic organization technology of mountain highway during incident state in adverse weather” (“Eleventh Five-Years” National Technology Supporting Program). Expressway were selected as the researched object. Two typical disasters, such as weather and emergent incidents, were selected as incident types. Based on the theory of operational risk analysis and the technology of traffic management, two operational safety levels of expressway including road sections and expressway network were studied in this article, which solved finally risk assessment problems of traffic flows from microscopic level and mesoscopic one during the typical incident state. Firstly, the characteristics of disasters weather, such as rain, fog, ice and snow etc., were analyzed, which were combined with each other as different kinds of incident. Based on impact caused by fog on the driver's sight distance, Fog-caused Driving Delay Time Model were put forward, which could quantify influence on the driver's reaction time in different visibility range. Based on the traditional theory of vehicle operation, numerical simulation of vehicle anti-yaw and anti- roll in different weather conditions were analyzed. At the same time, considering typical emergent incidents result caused by different kinds of incidents, such as traffic accidents, traffic events, natural disasters etc., the risk grades of expressway were established. Secondly, three typical operating modes on the expressway, such as free operation, car-following operation, lane-changing operation etc., were analyzed. Based on the decreasing impacts of visibility and road surface caused by adverse weather, different safety constraint condiction were calculated and safety operating standards were founded. Based on vehicle trajectory characteristics named as continuous reverse curve, geometric descriptic model of lane-changing was proposed firstly, which integrated classic car-following safety model into perfect safety theory. Thirdly, based on the characteristics of dynamic systems and risk methods, two risk situations including continuous traffic flow and blocking traffic flow were selected to be analyzed. The former was used to analyze small probability incidents which would cause serious results. On the contrary, the latter was used to analyze large probability incidents which would cause minor results. Considering damages to human in different driving conditions, different risk levels of were established. On these basis, technology and standards about detecting traffic risk were proposed. Fourthly, based on realistic traffic characteristics of anisotropy and heterogeneity, dynamic-classified method was proposed, which could be used to determine whether traffic flow interated or not. Two principles including direct-related classified and indirect-related classified were developed to discriminate interations between vehicles in road section. By constructing interrelated matrix and risk matrix, operating risk index of continuous traffic flow were founded, which could used to evaluate real-time traffic risk of continuous flow. Based on the relation between road supply and traffic demand, two indicators including average headway and cumulate flow rate were selected to evaluate operational risk of traffic flow during lane block. Finally, according to safety level management and goals of expressway network, meso-operations risk indicators named as Dynamic Risk Saturation, was proposed on the basis of micro-risks, which could be used to evaluate the operation of network-level safety during incidents. Considering varying characteristics of time and space, network equilibrium model were founded to satisfy needs of road users and management users, which evacuated traffic flows rapidly after satisfying operational safety constraints at every step of traffic assignment. After that, traffic safety measures, such as limiting operating speed and lane using, could be developed eventually. Selecting the Jing-zhu expressway network in Henan province, network safety technology were introduced to analyze normal state and incident state. In a word, operational risk analysis techniques of traffic flow were carried out from the two levels of road sections and network during typical incidents. The risk analysis methods based on traffic flows characteristics were completed primitively, which could provide a theoretical basis for expressway safety management. Key Words: expressway, disaster incident, continuous traffic flow, blocking traffic flow, operating safety, risk estimate, dynamic-classified method, lane-changing safety model, dynamic risk saturation, operational safety balance. |
博士论文