【作者】 刘黎；

【导师】 胡海波；

【作者基本信息】 重庆大学 ， 软件工程， 2021， 硕士

【摘要】 当今,随着人们出行需求的不断增加,城市公共交通工具(公交和地铁)的调度和运力已经不能很好地满足人们日常所需。网约车因其方便、快捷、经济等特点,承载了大量出行需求,成为城市交通工具的重要组成部分。网约车运行过程中会产生大量的运行数据,即订单数据和轨迹数据。订单数据包含丰富的出发点-目的地信息,轨迹数据包含丰富的道路交通信息,这些信息对于研究乘客出行行为、道路的流量演变和拥堵等城市交通问题以及优化城市交通有着重要意义。可视分析作为一种分析数据、发现问题的重要方法,目标在于通过多种图形图表,利用多种交互,准确易懂地表现数据内在规律,实现数据的层次化展示。针对现有网约车订单数据和轨迹数据分析方法在分析成本、分析效率、可理解性等方面存在的不足,本文提出基于信息可视化理论和技术的网约车运行数据探索与分析方法,同时设计实现出交互式的可视分析系统来支撑分析工作。本文的主要工作可以具体包括:(1)在可视化视图设计方面,本文提出了一种新型的订单雷达图。该视图将订单起点和终点的具体空间位置抽象出来,结合距离和方向属性,呈现起点和终点的时空概览。(2)在可视分析方法方面,订单雷达图可作为订单数据分析的交互入口,与地图、桑基图、折线图等多种组件实现联动,数据分析用户可结合时间、空间、语义三个角度,交互式地分析乘客的出行行为以及关键位置的热度变化。综合使用地图、像素图、时间线、词云图等经典视图组件,对轨迹数据进行多维度呈现和协作式分析,支持数据分析用户探索道路交通流量的演变、拥堵及其成因。(3)在可视分析系统方面,本文结合数据处理流程和可视分析任务,设计实现了一个多视图交互的可视分析原型系统。该系统注重各视图模块间的协同,通过丰富的交互将各视图关联,允许用户根据自己兴趣去探索整体与细节层面的信息。(4)在案例研究方面,本文采用滴滴公司发布的网约车订单数据和轨迹数据,结合实际应用场景,对视图设计的合理性、视图交互的有效性等方面验证与评估,并与业界的相关工作进行了分析比较。更多还原

【Abstract】 With the continuous increasement of people’s travel demand,the scheduling and transport capacity of urban public transportation(bus and subway)often can not meet people’s daily needs well.Due to its features of convenience,speed and economy,online car-hailing has carried a large number of travel demands and become an important part of urban transportation.During the operation of online car-hailing,a large amount of operation data will be generated,namely order data and trajectory data.Order data contains a wealth of origin and destination information,and trajectory data contains a wealth of road traffic information,which is of great significance to the study of passenger travel behavior,road flow evolution and congestion and other urban traffic problems,as well as the optimization of urban traffic.Visual analysis,as an important method to analyze data and find problems,aims to display the inherent laws of data accurately and easily through a variety of graphs and charts,using a variety of interactions,so as to realize the hierarchical display of data.In view of the shortcomings of the existing vehicle operation data analysis methods in the analysis of cost,analysis efficiency and comprehensibility,this paper proposes to use visual analysis to realize the exploration and analysis of the trajectory data of online car-hailing orders.An interactive visual prototype system is designed to support the analysis.The main work of this paper can be summarized as follows:(1)In the aspect of visual view design,this paper presents a new order radar chart.This view abstracts the specific spatial position of the vehicle origin and destination points,and presents a spatio-temporal overview of the origin and destination points by combining the distance and direction attributes.(2)In terms of visual analysis methods,the order radar chart can be used as an interactive entry for order data analysis,which can be linked with maps,sankey charts,line diagrams and other components.Data analysis users can interactively analyze the travel behavior of passengers and the heat changes of key locations from three perspectives,including time,space and semantics.Classic view components such as map,pixel map,time line and word cloud map are comprehensively used to present multi-dimensional and collaborative analysis of spatio-temporal data of vehicle tracks,supporting data analysis for users to explore the evolution of urban traffic flow,traffic congestion and its causes.(3)In the aspect of visual analysis system,this paper designs and implements a multi-view interactive visual analysis prototype system by combining data processing flow and visual analysis task.The system focuses on the collaboration between the view modules,connects the views through rich interactions,and allows users to explore the overall and detailed information according to their own interests.(4)In terms of case study,this paper uses the online ride-hailing order data and trajectory data released by Didi company,combined with the actual application scenarios,verifies and evaluates the rationality of view design,and the effectiveness of view interaction,and makes an analysis and comparison with the relevant work in the industry.更多还原