基于时限性的疏散路网扩容重点动态位置识别研究

It plays an important role to increase the capacity of some links or intersections for improving the total evacuation capacity of a road network in regional evacuation traffic management, in which the identification of vital positions for capacity-increasing is a critical technical problem. Considering the available time or period of network facilities, the project turns to study the problem of indentifying the vital dynamic locations or parts in a road network for increasing the total evacuation capacity, using modified critical edge and modified critical node models. To reflect the influence of the dynamic variation of some properties of a road network and the directional capacity of intersections, an evacuation road network will be abstracted as a special dynamic network with directional weights at nodes. The problem of critical dynamic edge or turning-direction for improving the universal maximal flow in such network will be defined and solved to indentify the vital dynamic links or turning lanes for capacity-increasing management, a dynamic maximal capacity path based heuristic algorithm will be presented. In addition, a bi-level programming model will be presented to indentify the vital combination of dynamic links and intersection turning lanes in the case of comprehensive capacity-increasing management. The upper level uses a multi-objective programming model to optimize the scale of capacity-increasing (the total number of dynamic links and turning lanes involved in capacity-increasing). The lower level indentifies the vital dynamic combination of links and turning lanes under a given scope by searching the critical dynamic combination of edges and turning-directions for improving the universal maximal flow. At last, an evacuation planning decision aided system will be developed by integrating evacuation traffic flow simulation and the study results mentioned above through programming tools such as c++, which can tell the effect of capacity-increasing management and adjust capacity-increasing plans.

大面积疏散中，对部分路段或交叉口进行“扩容”管理，是提高路网整体疏散通行能力的重要手段，而扩容重点位置识别是其中的关键技术问题。本项目考虑道路设施的可用时限，通过改进经典的关键边、关键点模型，研究疏散路网扩容重点动态位置的识别问题。为反映道路设施性能的动态变化及交叉口方向性通行能力的影响，将路网抽象为动态方向性点权网络。通过定义并求解该类点权网络中的全局最大流增流关键动态边/转向问题，识别重点扩容重点动态路段或转向，设计基于动态最大容量路的启发式算法。进一步考虑路段与交叉口的组合扩容，采用双层规划模型对扩容重点动态路段及转向组合进行选择，其中上层通过多目标规划确定扩容规模，下层则对给定的规模，通过在该类点权网络中寻找全局最大流增流关键动态组合，识别扩容重点动态组合。最后，通过c++等编程工具，将疏散交通流仿真与以上研究结果集成，开发疏散规划辅助决策系统，实现扩容管理方案的验证及反馈调整。

大面积疏散中，对路网中部分路段或交叉口进行“扩容”管理，是提高路网整体疏散通行能力的重要手段。在有限的应急资源和执行力条件下，有针对性地识别和选择需要扩容的重点位置是扩容管理中的关键技术问题。本项目基于网络流理论，结合扩容重点位置识别的具体含义，首先定义了最大流增流关键边问题，给出了相应模型和算法，将之应用于扩容关键路段的识别问题。然后在此基础上进行改进，给出了方向性点权网络中最大流增流关键边问题得模型和算法，用于在考虑交叉口影响时扩容关键路段的选择优化，实例表明考虑交叉口影响所得的扩容方案更加合理。接着提出了全局最大流增流动态关键边的概念并设计了相应的求解算法，用之进行了最佳扩容路段与时限的选择优化问题，实例结果表明，动态扩容方案较之静态扩容方案，能更有效地压缩疏散时间。在全局最大流增流动态关键边模型的基础上进一步改进，给出了在方向性点权网络中寻找全局最大流增流动态关键边的模型和算法，用于在考虑交叉口影响的情况下，识别扩容重点路段及时限，实例结果表明该种方案更有针对性。另外定义了方向性点权网络中的全局最大流增流关键组合（路段与交叉口转向组合）问题，在前述问题所用的动态最小费用路算法基础上进行改进，给出了寻找关键组合的有效算法，其中通过记录每个增广路中的边和转向以提高搜索效率，计算结果即关键组合中对应的边和转向可作为应急疏散组织中路段和交叉口交通管制的依据。最后将上述模型和算法融合，结合兰州市西固区调研数据，开发了面向该区域的疏散规划辅助决策系统，使重工业危险源区域的应急交通组织更具有针对性，同时让普通的应急管理人员也能使用项目研究成果。