基于MFD的城市交通路网控制子区域划分与子区域边界交叉口重要度评估方法研究

发布时间：2018-10-12 11:09

【摘要】：科技的发展,社会的进步,使得大中型城市的人口急剧膨胀,城市居民的出行的需求剧增。同时电商的崛起加快了物流运输业的发展,使交通服务愈加供不应求。大中型城市交通路网拥堵成为全球问题。近年来,国外一些专家学者不再聚焦于单个拥堵交叉口控制策略的研究,而是从整个交通路网的基本属性——宏观基本图(MFD)出发,将整个城市路网进行子区划分(划分后的各个子区具有同质MFD特征),对相邻子区域之间采取边界控制方法,控制区域间输入和输出交通流。实践证明,这是解决交通拥堵的一种有效方法。我国关于MFD的研究刚刚起步。如何在已有城市交通基础设施的基础上,通过探究城市路网的MFD特性,进行城市路网控制子区域划分,设计有效的控制策略,成为解决城市交通拥堵的一个前沿课题。本文基于北京市望京地区浮动车的实际数据,提出了基于MFD的网格逐层的城市路网控制子区域划分的新方法。同时从高效节能的角度考虑,本文提出仅对子区域边界上的关键节点进行边界控制。充分考虑一般复杂网络的基本特性和实际路网的拓扑结构,提出一种切实可行的交通路网交叉口重要度的评价方法——基于m阶邻居节点重要度的评估算法,并通过仿真分析,与已有方法进行分析和比较,验证了该算法具有较好的可行性和精准性。为下一步实现部分路网节点的边界控制提供理论依据。首先,本文对采集的北京市望京地区浮动车数据进行预处理。以高德地图为基础,绘制出其路网结构图,统计各个路段信息。根据采集到GPS的车辆定位数据,利用投影算法与道路进行匹配。依据采集到的浮动车的瞬时速度、车辆标识、车辆状态等数据计算出本文所需交通参量,如:车流量、平均速度、平均密度等。其次,利用所求交通参数,研究整体路网的MFD特性;对研究区域进行分析,采用网格逐层试凑叠加的方式进行区域划分,并以城市路网的MFD(宏观基本图)特性,如:均匀性、离散度低、滞环现象等,作为一系列的评价指标用于评价区域划分的结果。最后,对已有的一般复杂网络关键节点选取评价方法和相关的道路交通指标进行了分析,选择在实际路网中体现比较好的特性,如度值、繁忙度、紧密度中心性及介数中心性等,并在此基础上提出一种新的基于m阶邻居节点重要度的关键交叉口评估方法,建立新的路网中关键交叉口的综合评价指标。采用基于蚁群算法的BP神经网络进行评估过程中的权值优化。与多种评价指标进行了对比和评估,表明本文方法的可行性和有效性。
[Abstract]:With the development of science and technology and the progress of society, the population of large and medium-sized cities expands rapidly, and the demand of urban residents to travel increases dramatically. At the same time, the rise of e-commerce accelerated the development of logistics industry, so that transport services more than demand. Traffic congestion in large and medium-sized cities has become a global problem. In recent years, some foreign experts and scholars no longer focus on the control strategy of a single congested intersection, but start from the basic attribute of the whole traffic network, which is the macroscopic basic diagram (MFD). The whole urban road network is divided into sub-areas (each sub-area has homogeneous MFD characteristics) and the boundary control method is adopted to control the input and output traffic flow between the adjacent sub-regions. Practice has proved that this is an effective method to solve traffic congestion. The research on MFD in China has just started. On the basis of the existing urban traffic infrastructure and by exploring the MFD characteristics of the urban road network, how to divide the control sub-area of the urban road network and design effective control strategies has become a frontier problem to solve the urban traffic congestion. Based on the actual data of floating vehicles in Wangjing area of Beijing, this paper presents a new method of dividing the control subareas of urban road network based on MFD. At the same time, from the point of view of high efficiency and energy saving, this paper presents boundary control only for the key nodes on the sub-region boundary. Considering the basic characteristics of general complex network and the topological structure of practical road network, a feasible evaluation method of intersection importance of traffic network is proposed, which is based on the evaluation algorithm of m-order neighbor node importance, and is analyzed by simulation. Compared with the existing methods, the proposed algorithm is proved to be feasible and accurate. It provides a theoretical basis for the next step to realize the boundary control of some road network nodes. Firstly, this paper preprocesses the floating vehicle data collected in Wangjing area of Beijing. Based on Amap, draw out its road network structure map, statistics each section information. According to the vehicle location data collected from GPS, the projection algorithm is used to match the road. Based on the collected data of floating vehicle's instantaneous speed, vehicle identification and vehicle status, the traffic parameters are calculated, such as: traffic flow, average speed, average density and so on. Secondly, the MFD characteristics of the whole road network are studied by using the traffic parameters sought, and the study area is analyzed, and the area is divided by the method of grid layer by layer superposition, and the MFD (macroscopic basic map) characteristics of the urban road network, such as uniformity, are used to divide the area. As a series of evaluation indexes, low dispersion and hysteresis are used to evaluate the results of regional division. Finally, the paper analyzes the existing methods of selecting and evaluating key nodes in general complex network and the related road traffic indexes, and chooses the better characteristics in the actual road network, such as degree value, busy degree, etc. On the basis of compactness centrality and meso-centrality, a new evaluation method for key intersections based on the importance of m-order neighbor nodes is proposed, and a new comprehensive evaluation index of key intersections in road network is established. The BP neural network based on ant colony algorithm is used to optimize the weights in the evaluation process. Compared with many evaluation indexes, the method is feasible and effective.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491

【参考文献】