面向路网运行管理的异构传感网协同设计方法研究

发布时间：2018-07-31 16:46

【摘要】：高速公路网是支撑国民经济发展、服务百姓出行、保障国防安全的战略基础设施,截至2014年底,我国高速公路通车总里程接近11.2万公里,基本达到了规划的高速公路网规模,承担了全国公路30%以上的客货周转量。与此同时,百姓安全、便捷机动化出行体验、跨区域出行服务、极端恶劣气象频发等对大范围路网的管理与服务提出了更高的要求,从建设管理到面向路网运行监测、应急处置与服务的路网运行管理成为了必然选择。为实现路网运行管理的功能,发挥区域路网的综合运输潜力、改善跨区域大范围服务体验并提升突发状态的协同处置能力,路网运行管理部门需要:1)做好监测体系规划,可靠地估计和预测网络范围动态交通状态;2)做好路网管控体系设计,应急情况下,实现"远程疏导、近处分流、关键点段现场控制"的协同调度策略;3)综合利用车载和路侧信息服务设施,提供高效信息服务。由此可见,更好地适应区域路网管控、应急处置和公众服务的传感网的设计成为当前亟待解决的问题。与此同时,随着信息通信技术的发展,路网运行监测方面对断面交通流检测设备的依赖程度有所降低。随着全国高速公路联网不停车收费系统的建成,以专用短程通信技术(Dedicated Short Range Communication,DSRC)为代表的高速公路车路交互平台得到了快速发展;随着无人机技术的日趋成熟和低空领域的逐渐开放,以无人机为代表的移动传感技术在交通管理中取得了一定的应用。本论文研究的主要目的是以路网运行管理需求为切入点,融合新一代的交通运行监测技术,协同布设多源异构传感器,为路网协同管控、公众出行服务提供更可靠的信息。本论文以最优化理论和交通分配理论为基础,以路网管控和服务能力优化提升为目标,研究了基于均衡配流的关键路段辨识方法,提出了路网可观可控性的DSRC网最优化设计方法,构建了面向网络运行状态检测的无人机动态路线规划模型。主要研究成果包括:(1)开展多源异构传感装备的适用性分析。分析路网运行管理主要需求及监测要素,分析多源异构监测装备检测参数类型、数据质量、成本,及其对路网运行管理需求的适用性,为传感器的布局优化及多源异构传感网的协同设计奠定基础。(2)开展基于均衡配流的路网关键路段集辨识研究。综合考虑路网结构、交通需求、出行行为特征等多方面因素,针对多路段失效的联合效应,提出了基于K-短路径和用户均衡配流的公路网关键路段评估方法,构建了路段失效概率分布的关键路段集辨识模型;进一步针对路段抗毁性,以救援恢复成本为约束,构建了基于路段能力可恢复性的关键路段集辨识模型;并在不同网络上验证了模型的有效性和可行性。(3)开展面向路网可观可控性的DSRC网最优化设计方法研究。具有双向通信特征的DSRC路侧设备,可同步实现交通运行状态信息的获取和诱导信息的发布功能。本论文针对高速公路网络化监测与管控需求,提出了路网可观性、可控性框架模型;以建设成本为约束条件,构建了最小化路网旅行时间不确定性及最小化系统总旅行时间为目标的可观、可控性优化模型,实现路网范围内DSRC网络的优化布局;利用高速公路通道对模型进行了验证。(4)开展面向网络运行状态检测的无人机动态路线规划模型研究。针对突发事件快速监测以及日常巡查需求,考虑路网的物理拓扑结构以及事件的时空扩散性,本论文引入时空地理框架,以最小化无人机和固定传感器在时空网络上的事件延误检测成本为目标,以无人机飞行成本和固定传感器建设费用为约束,提出了融合固定传感器的无人机路线动态规划模型;针对固定交通流传感器已知和未知两种情况,分别构建了拉格朗日松弛算法,其中,固定交通流传感器已知情况下,将模型简化为经典的时变最小路径成本问题;利用不同级别的路网进行了模型和算法验证。
[Abstract]:The expressway network is a strategic infrastructure to support the development of the national economy, to serve the people and to ensure the security of national defense. By the end of 2014, the total mileage of the highway in China is close to 112 thousand kilometers, basically reaching the planned highway network scale and taking over 30% of the passenger and cargo turnover of the national highway. At the same time, the people are safe and convenient. Mobile travel experience, cross regional travel service, extreme weather frequency and so on have put forward higher requirements for the management and service of large-scale road network. From construction management to road network operation monitoring, emergency disposal and service management of road network has become an inevitable choice. Comprehensive transportation potential, improving the service experience across the region and improving the cooperative disposal capacity of the sudden state, the road network operation and management departments need: 1) to plan the monitoring system, reliably estimate and predict the dynamic traffic state of the network range; 2) to set up the network management and control system, and to realize the "remote guidance, the near diversion, the key to the key" in the emergency situation. Point field control "cooperative scheduling strategy"; 3) comprehensive use of vehicle and road side information service facilities to provide efficient information services. Thus, it can be seen that better adaptation to regional network management, emergency disposal and public service sensing network design has become an urgent problem. At the same time, along with the development of information and communication technology, road network transportation The dependence degree on cross section traffic flow detection equipment is reduced. With the completion of the national freeway network, the expressway interactive platform, represented by Dedicated Short Range Communication (DSRC), has been developed rapidly; with the increasing number of unmanned aerial vehicles (UAV) technology With the gradual opening up of the mature and low altitude fields, the mobile sensing technology represented by unmanned aerial vehicles (UAV) has been applied in traffic management. The main purpose of this paper is to integrate the new generation of traffic monitoring technology with a new generation of traffic monitoring technology, and to coordinate the network management and control of the road network. On the basis of optimization theory and traffic distribution theory, this paper studies the key link identification method based on equilibrium distribution based on the optimization theory and traffic distribution theory, and puts forward the optimal design method of DSRC network with considerable controllability in road network, and constructs a network oriented operation form. The main research results include: (1) carry out the applicability analysis of multi-source heterogeneous sensing equipment, analyze the main requirements and monitoring factors of road network operation and management, analyze the types of detection parameters, data quality, cost, and the applicability of the requirements for the operation and management of the road network. The layout optimization of the device and the collaborative design of multi-source heterogeneous sensing network are laid. (2) to carry out the identification research on the key section of road network based on balanced distribution, considering many factors, such as road network structure, traffic demand, travel behavior characteristics and so on. In view of the joint effect of multi section failure, the K- short-circuit diameter and the user equilibrium distribution are proposed. The key section evaluation method of road network key section has constructed the key link set identification model of the link failure probability distribution. Further, the key section identification model based on the recoverability of section capacity is constructed for the road damage resistance and the recovery cost of rescue is constrained, and the validity and feasibility of the model are verified on different networks. (3) The optimization design method of DSRC network oriented to the observable controllability of road network. The DSRC roadside device with two-way communication features can synchronize the acquisition of traffic state information and the release function of the induced information. In this paper, the observability and controllability frame model of road network is proposed for the network monitoring and control requirements of the expressway. In order to minimize the uncertainty of road network travel time and minimize the total travel time of the system, the optimal layout of the DSRC network within the network range is realized, and the model is verified with the highway channel. (4) Unmanned maneuver for network running state detection is carried out. In order to minimize the cost of event delay detection of unmanned aerial vehicles and fixed sensors on the space-time network, this paper aims at the rapid monitoring of emergency events and the demand for daily inspection, considering the physical topology structure of the road network and the space-time diffusivity of events. Based on the construction cost of fixed sensor, a dynamic programming model of unmanned aerial vehicle (UAV) with fixed sensor is proposed. According to two known and unknown conditions of fixed traffic flow sensor, the Lagrange relaxation algorithm is constructed respectively. In the case of fixed traffic flow sensor, the model is simplified as the classic time varying minimum path. The cost and the model and algorithm are verified by using different levels of road network.
【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TP212.9;U495

【相似文献】