城市道路动态交通流仿真的研究与实践
发布时间：2018-11-04 13:25
【摘要】：交通流仿真是智能交通运输系统的一个重要组成部分,它可以动态地、逼真地仿真交通现象,展现交通流的时空变化,为交通规划、交通组织与管理、物资运输流量优化等方面的研究提供仿真测试平台。然而,仿真结果与实际运行状况之间往往存在较大的差异,影响了交通流仿真的有效性和实用性。本文以提升交通流仿真的真实性为目标,从路网建模、OD (Orignation Destination)预测、车辆路径选择等方面展开研究,提出并实现了相应的改进方法,探讨了交通流实测数据对仿真模型的修正机制。 本文首先研究并实现了交通流仿真系统的静态物理模型的构建,主要包括城市道路模型的构建、车辆行驶路线的设置、仿真实体车辆模型的构建和模型间的碰撞检测等；研究了城市时变出行OD矩阵的获取方法,提出并实现了基于历史数据和重力模型的出行分布时变特性的预测,可以减少OD调查工作量,也能较客观地反映实际分布。 在动态交通流分配方面,本文分析了现有路阻函数模型的不足,提出了更符合国内实际交通情况又易于计算的改进模型；根据实际道路中驾驶员的出行行为特性,提出基于多项式分对数(MNL)模型的车辆动态路径选择方法来实现交通流的生成,较好地表达了实际交通流形成过程的动态性和随机性；根据交通需求和交通路网特性,提出车辆出发时间模型、有效路段的判别以及有效路径的计算方法。在仿真模型的校准机制方面,结合交通检测器的实测数据和交通流的仿真模型,利用卡尔曼滤波器对交通仿真状态进行实时预测校正。 本文最后分析并实现了交通流仿真原型系统的总体结构,以某一具体路网为例,依次实现了车辆的生成、交叉口前的路径选择和路段上的运动仿真,出行车辆的这些一系列行为构成了城市道路中的宏观交通流。从仿真结果图可以看出驾驶员动态路径选择所形成的交通流分布符合城市道路的交通流特点,证明了路径选择模型的可行性和合理性。
[Abstract]:Traffic flow simulation is an important part of intelligent transportation system. It can dynamically and realistically simulate traffic phenomena, show the temporal and spatial changes of traffic flow, and make traffic planning, traffic organization and management. The research of material transportation flow optimization provides a simulation test platform. However, there is often a big difference between the simulation results and the actual operation status, which affects the effectiveness and practicability of traffic flow simulation. In order to improve the authenticity of traffic flow simulation, this paper studies the road network modeling, OD (Orignation Destination) prediction, vehicle routing and so on, and puts forward and realizes the corresponding improvement method. The modification mechanism of the simulation model based on the measured data of traffic flow is discussed. Firstly, the static physical model of the traffic flow simulation system is studied and realized in this paper, including the construction of the urban road model, the setting of the vehicle route, the construction of the simulation entity vehicle model and the collision detection between the models. In this paper, the acquisition method of OD matrix for urban time-varying travel is studied, and the prediction of time-varying characteristics of travel distribution based on historical data and gravity model is proposed and realized, which can reduce the workload of OD survey and objectively reflect the actual distribution. In the aspect of dynamic traffic flow distribution, this paper analyzes the deficiency of the existing road resistance function model, and puts forward an improved model which is more in line with the actual traffic situation in China and is easy to calculate. According to the characteristics of the driver's travel behavior in the actual road, a method of vehicle dynamic path selection based on the polynomial divided logarithmic (MNL) model is proposed to generate the traffic flow. The dynamic and randomness of the formation process of actual traffic flow is well expressed. According to the traffic demand and the characteristics of traffic network, the vehicle departure time model, the discrimination of effective section and the calculation method of effective route are proposed. In the aspect of the calibration mechanism of the simulation model, combining the measured data of the traffic detector and the simulation model of the traffic flow, the Kalman filter is used to predict and correct the traffic simulation state in real time. Finally, the overall structure of the prototype system of traffic flow simulation is analyzed and realized. Taking a specific road network as an example, the generation of vehicles, the path selection before intersections and the motion simulation on road sections are realized in turn. These behaviors of vehicles constitute the macro traffic flow in urban roads. From the simulation results, it can be seen that the traffic flow distribution formed by the driver's dynamic path selection accords with the characteristics of the urban road traffic flow, which proves the feasibility and rationality of the route selection model.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.112
本文编号：2309947
[Abstract]:Traffic flow simulation is an important part of intelligent transportation system. It can dynamically and realistically simulate traffic phenomena, show the temporal and spatial changes of traffic flow, and make traffic planning, traffic organization and management. The research of material transportation flow optimization provides a simulation test platform. However, there is often a big difference between the simulation results and the actual operation status, which affects the effectiveness and practicability of traffic flow simulation. In order to improve the authenticity of traffic flow simulation, this paper studies the road network modeling, OD (Orignation Destination) prediction, vehicle routing and so on, and puts forward and realizes the corresponding improvement method. The modification mechanism of the simulation model based on the measured data of traffic flow is discussed. Firstly, the static physical model of the traffic flow simulation system is studied and realized in this paper, including the construction of the urban road model, the setting of the vehicle route, the construction of the simulation entity vehicle model and the collision detection between the models. In this paper, the acquisition method of OD matrix for urban time-varying travel is studied, and the prediction of time-varying characteristics of travel distribution based on historical data and gravity model is proposed and realized, which can reduce the workload of OD survey and objectively reflect the actual distribution. In the aspect of dynamic traffic flow distribution, this paper analyzes the deficiency of the existing road resistance function model, and puts forward an improved model which is more in line with the actual traffic situation in China and is easy to calculate. According to the characteristics of the driver's travel behavior in the actual road, a method of vehicle dynamic path selection based on the polynomial divided logarithmic (MNL) model is proposed to generate the traffic flow. The dynamic and randomness of the formation process of actual traffic flow is well expressed. According to the traffic demand and the characteristics of traffic network, the vehicle departure time model, the discrimination of effective section and the calculation method of effective route are proposed. In the aspect of the calibration mechanism of the simulation model, combining the measured data of the traffic detector and the simulation model of the traffic flow, the Kalman filter is used to predict and correct the traffic simulation state in real time. Finally, the overall structure of the prototype system of traffic flow simulation is analyzed and realized. Taking a specific road network as an example, the generation of vehicles, the path selection before intersections and the motion simulation on road sections are realized in turn. These behaviors of vehicles constitute the macro traffic flow in urban roads. From the simulation results, it can be seen that the traffic flow distribution formed by the driver's dynamic path selection accords with the characteristics of the urban road traffic flow, which proves the feasibility and rationality of the route selection model.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.112
【参考文献】
相关期刊论文 前10条
1 陈尚云,杜文,高世廉;我国特大城市出行分布模型及其参数的研究[J];系统工程;2002年04期
2 袁振洲;基于计算机模拟的动态随机分配优化算法的研究[J];公路交通科技;2002年01期
3 何勇;我国道路交通安全现状及对策[J];公路交通科技;2003年S1期
4 周溪召;驾驶员实时动态路径选择行为组合模型[J];上海理工大学学报;2002年02期
5 邹智军,杨东援;道路交通仿真研究综述[J];交通运输工程学报;2001年02期
6 周光伟,杨晓光;城市道路交叉口车辆微观行驶模型[J];交通运输工程学报;2001年04期
7 张兰;彭国雄;;重力模型标定方法的分析及应用[J];交通科技与经济;2009年01期
8 段进宇,缪立新,江见鲸;由路段交通流量反估出行OD矩阵技术的应用[J];清华大学学报(自然科学版);2000年06期
9 袁雄军;张华;;我国道路交通安全的现状及对策分析[J];交通与运输;2008年05期
10 褚琴,陈绍宽;重力模型标定方法及应用研究[J];交通运输系统工程与信息;2003年02期
，本文编号：2309947
本文链接：https://www.wllwen.com/kejilunwen/jiaotonggongchenglunwen/2309947.html
