基于感知数据的交通仿真研究

发布时间：2018-02-04 23:57

本文关键词： 交通仿真感知数据电子足迹 matsim 体系结构　出处：《天津大学》2014年硕士论文　论文类型：学位论文

【摘要】：大量的对于现实世界交通的互动仿真能够给决策者在进行道路交通规划、建设的过程中提供极大的参考价值。然而要进行道路交通仿真,相应的仿真数据必不可少,如何廉价、方便、全面地获得仿真所需的相关数据并利用合适的方法框架挖掘其内部交通模式作为仿真输入是当前实施交通仿真亟需解决的一大难题。本文提出了基于感知数据的交通仿真方法框架,主要利用人类电子足迹数据依托大数据仿真平台进行交通模式的挖掘和交通的仿真,完成了以下工作:第一,提出了使用人类电子足迹数据进行交通参数以及相应交通模式挖掘的方法框架。本文将详细阐述计算交通相关参数的过程,包括交通流量和速度的计算,以及对车辆轨迹的挖掘和应用车辆轨迹进行OD分析、出行端点识别和出行计划分析的过程。第二,分析设计了大数据仿真平台体系结构。本文使用大数据仿真平台作为交通仿真实验的承载工具,对该平台的需求、架构做了详细的分析和设计,为交通仿真提供了平台依据。第三,通过大数据仿真平台结合matsim工具设计并完成交通仿真实验。本文首先使用已提出的方法框架结合大数据仿真平台对人类电子足迹数据和地理信息数据进行预处理,形成路网数据、建筑设施数据、出行计划数据和验证数据,将这些数据作为matsim道路仿真工具的输入进行仿真,然后将仿真的最终结果进行可视化的输出展现。人类电子足迹数据往往通过已有公共基础设施进行获取,因此该数据不仅具有高密度的人类行为信息,还有高覆盖、廉价和易于获取的特点。利用该数据进行交通参数和模式的挖掘进而依托大数据仿真平台形成仿真的输入数据并进行仿真可以使得仿真的过程得到简化,仿真的结果更加完备和精确,让仿真的执行者摆脱仿真前期数据获取和预处理的困扰,更多地集中于交通仿真本身。
[Abstract]:A large number of interactive simulation for real world traffic can provide great reference value for decision-makers in the process of road traffic planning and construction. However, road traffic simulation should be carried out. The corresponding simulation data is necessary, how cheap and convenient. It is a difficult problem for traffic simulation to obtain the relevant data and mine its internal traffic pattern as the simulation input by using the appropriate method framework. In this paper, we propose a method based on perceptual data to solve the problem of traffic simulation. Traffic simulation method framework. Mainly using human electronic footprint data based on big data simulation platform for traffic pattern mining and traffic simulation, completed the following work: first. A framework for mining traffic parameters and corresponding traffic patterns using human electronic footprint data is proposed. The process of calculating traffic related parameters, including the calculation of traffic flow and speed, is described in detail in this paper. And the vehicle trajectory mining and application vehicle trajectory OD analysis, travel endpoint identification and trip planning analysis process. Second. This paper uses big data simulation platform as the carrying tool of traffic simulation experiment, and makes a detailed analysis and design to the demand of the platform and the architecture. Provide the platform basis for traffic simulation. Third. The traffic simulation experiment is designed and completed by using big data simulation platform and matsim tools. Firstly, this paper uses the proposed method framework and big data simulation platform to study human electronic footprint data and geographic information data. Preconditioning. The road network data, building facilities data, trip planning data and validation data are formed and simulated as the input of matsim road simulation tools. Then the final results of the simulation are visualized. Human electronic footprint data are often obtained through the existing public infrastructure, so the data not only has high density of human behavior information. And high coverage. The data can be used to mine the traffic parameters and patterns to form the simulation input data based on the big data simulation platform and the simulation can simplify the simulation process. The simulation results are more complete and accurate, so that the simulation executors can get rid of the problems of data acquisition and preprocessing in the early stage of simulation, and focus more on the traffic simulation itself.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491;TP391.9

【相似文献】