《北京市交通大数据应用和展望ppt课件.pptx》由会员分享，可在线阅读，更多相关《北京市交通大数据应用和展望ppt课件.pptx（36页珍藏版）》请在三一办公上搜索。
1、北京交通大数据应用与展望Applications and Prospect of Beijing Transportation Big Data,汇报内容Content,2,A,B,C,数据类型 Data Type,基础设施 Infrastructure,车辆和人员Vehicles and Employees,交通行为数据Behavior Data,交通检测数据Detection Devices,人口、土地、经济Population, Land, Economics,气象、环境等Weather, Activities,数据规模 Data Scale,4,汇报内容Content,5,A,B,C,
2、应用1: 浮动车系统 (FCD),北京：40000辆出租汽车，5分钟（20秒）计算一次 Beijing: 40,000 taxis, 5 minutes (20 seconds) updating 五环内路网覆盖率80%，精度86%以上 Coverage (5th ring road): 80%, Accuracy: over 86%,Taxi GPS distribution,Real-time traffic congest,识别常发拥堵路段，并与土地开发关联分析，服务拥堵治理（2009-）,7,早高峰拥堵路段和节点,职住平衡度与早高峰常发拥堵路段,Congested segment an
3、d intersection Morning peak,Commercial/residential balance and congested segments morning peak,Identifies frequently congested road，Analysis with land development，Support for congestion management from 2009,8,Congestion formation and dissipation ratio - 2:3,Heavy rain, evening peak,Congestion Format
4、ion and Dissipation Evolution A Subtle Portrayal of Phantom,交通网络拥堵形成和消散演变规律分析 幽灵拥堵的细微刻画,大规模降雨、晚高峰,拥堵形成和消散的时间比例为2：3,9,Congestion Distribution Cloud Image,horizontal axis: time vertical axis: road segmentscolor : congestion level,An accident causing congestion.A: congestion spreadB: congestion dissipa
5、tion,Congestion Formation and Dissipation Evolution A Subtle Portrayal of Phantom,交通网络拥堵形成和消散演变规律分析 幽灵拥堵的细微刻画,交通拥堵分布时空“云图”,横轴代表时间，纵轴代表路段颜色代表该时间、该路段的拥堵程度。,莲花桥附近发生事故，产生拥堵。A：拥堵蔓延B：拥堵消散,10,Inspiration:,Can we use a figure to reflect transportation conditions?,Roads are like stocks,traffic index as its c
6、ore,Five-dimensional congestion recognition concept,应用二：“交通指数”拥堵评价Traffic Index Congestion Evaluation,灵感来源：,拥堵识别“五维”理念,能否用一个数字反映交通运行状态,道路类似个股,以“交通指数”为核心的拥堵评价体系,11,尾号4、9限行the last number on the license plate Is 4 or 9,very sensitive to changes in urban transport,交通指数对城市交通拥堵变化非常敏感,12,月交通指数变化（2007-2015
7、）Month traffic index （from 2007 to 2015）,数据积累：连续积累了2005年至今的所有数据。Data accumulation: continuously from 2005 to the present第一次定量化地衡量了城市拥堵，拥堵治理目标实现定量化。first time: urban congestion be measured quantitatively, and management goals also become quantitative.,severe congestion during evening peak,2010,72 day
8、s,2011,22 days,2012,28 days,2014,45 days,晚高峰严重拥堵天,14,Congestion duration （from 2007 to 2015）,serious,Light,middle,交通运行提前研判和应对 Congestion forecasting and replying,建立研判工作机制，年初、节假日前、高峰拥堵期前，针对性进行研判预测，引导各部门运力调配、公众出行,Application 3: Electronic Fee (IC) for Public Transport,16,Segment passengers of bus netw
9、ork,public transport operational parameters,应用三：公交电子收费（IC卡）,公交网络断面客流,17,Boarding and alighting volume,Congested segments and nodes,Bus passenger flow concentration locations generally have serious congestion,城市公交客流集中的地方，往往拥堵较为严重,基于公交IC卡数据计算 From bus IC data,基于浮动车行程速度计算 From floating car data,全市客流点登降
10、量,晚高峰拥堵路段和节点,18,Subway Passenger Characteristics and Land Use,Residential,Commercial,Residential+Commercial,轨道交通客流特征与周边用地功能分析,居住主导：天通苑,商业主导：国贸,居住+商业：军博,Taxi Passenger Characteristics Analysis,19,应用四：出租汽车客流特征分析,早高峰7:00-9:00,晚高峰17:00-19:00,出行速度分布（按起点位置统计）,上地、中关村,二环内,西四环,早高峰：二环内南部、上地、中关村，出行速度较低；,晚高峰：三环
11、内、北部城区（中关村、上地、望京、东北四环），低速,20,出行距离(单位：公里),四环内,四环-五环,五环外,8.1公里,9.9公里,12.6公里,特点：内短外长、西短东长,白天6:00-22:00,出行距离分布（按起点位置统计）,Taxi Passenger Characteristics Analysis,应用四：出租汽车客流特征分析,21,直达系数(直达系数=直线距离/行驶距离),棋盘状路网，理想直线系数=1.414/2=0.707,回龙观、天通苑,南四环外,直达系数：0.66,直达系数：0.67,回龙观,南四环外,Taxi Passenger Characteristics Analy
12、sis,应用四：出租汽车客流特征分析,Application 5: Analysis Based on Pseudo-code mobile signaling Data,22,About 18 million users a day (sample), 1 billion records everyday,应用五：伪码移动信令数据的交通运行分析,每日接近1800万用户（样本），每天10亿条记录。,在京外地人口中，前三甲：河北（25%）、河南（10%），山西（8.94%）,在京外地人口中（34天持续监测数据）在京停留时间超过27天的：占54%在京停留时间 2-7 天的：占总量的13%,Iden
13、tifies the population (mobile phone users) relevant to the CBD Computes travel statistics,24,Travel Characteristics Analysis of CBD Population,Average Daily Attraction Total,CBD population 24-hour movement,2-4时,4-6时,6-8时,8-10时,10-12时,12-14时,14-16时,16-18时,18-20时,20-22 h,CBD区域出行特征分析,从全市样本数据中，识别出与CBD相关
14、的人口(伪码移动信令用户)，针对这部分群体，统计其出行行为。,日均吸引总量,CBD相关人口24小时流动,Analysis of Public Rail Passenger Characteristics,25,Beijing subway hourly sectional flow diagram for morning peak,From bus IC Data,From Pseudo-code mobile signaling Data,轨道交通乘客特征分析,由公交IC卡数据计算,北京轨道交通早高峰小时断面流量图,由伪码移动信令数据计算,地铁1号线乘客的居住地分布（人口数）,地铁1号线乘客
15、的工作地分布(人口数),Urban Transport Simulation Model,26,应用六：城市交通仿真模型,27,Jingtong Expressway Bus Lane Deployment Case,Sectional bus passengers distribution (morningpeak),Bus routes are concentrated, the traffic is very heavy so buses and private vehicles mutually affect each other, drop efficiency.,From IC d
16、ata,Tongzhou to CBD passenger distribution,From Pseudo-code mobile signaling Data,京通快速路设置公交专用道案例,市政府拟设置公交专用道，实现公交和私人交通双赢。,京通快速路沿线公交线路密集，但交通负荷度高，公共交通和私人机动车相互影响，效率均较低。,Proposed to deploy dedicated bus lanes to benefit both public and private transport.,京通快速路断面公交客流OD分布（早高峰）,通州至CBD客流分布,28,Bus Lane Deplo
17、yment Simulation,Morning peak,Evening Peak,公交专用道设置仿真分析,测试结果：早高峰进城方向、晚高峰出城方向减少一条车道，交通需求会有所下降，路段负荷有所增加，周边平行道路流量有所增长，但总体效果能够有效缓解京通快速路拥堵情况，实施方案可行。,Result Evaluation (Open for 1 Year),高峰时段公交运力提高35% Capacity（peak）: + 35%,公交IC卡数据 From bus IC data,浮动车速度数据 From Floating Car Data,About 25% of passengers from
18、subway to bus,开通前,开通后,before,after,开通前 before,开通后 after,效果评估（开通1年后）,全日客运量增长24.5% passenger s（day）：+ 24.5%,乘客问卷调查数据 From Passenger Survey Data,约25%的乘客由地铁转移来,30,A,B,C,汇报内容Content,Mobile Internet: more data, complex association,31,公交、自行车、私家车 Bus, bicycle, car客运与货运 Passenger and freight交通与土地、人口 Transpor
19、tation and land, population 出行与社交、餐饮 Travel and social, catering,移动互联：数据将越来越多，且复杂关联,如浮动车：4万辆 = 500万辆 FCD： 40 thousand = 5 million 出行意愿：12万 = 2100万人 SP：120 thousand persons=21 million,完整的出行链路Completetravel link起点、换乘点、终点Start, transfer, end步行、公交车、bicycle、bus, ,空间： 10 米 = 0.1米Special：10 meter = 0.1 met
20、er时间：60秒 = 1秒Interval ：60second= 1 second时效：实时time：real time,上班、公务、下班、娱乐工作日、休息日、节假日今年、去年、过去十年,Influencefor urban, transportation planning,供给和需求如何平衡？How to balance supply and demand出行者需求：用户最优 Traveler demand: the user optimal网络供给：系统最优 Network supply: the system optimal充分互联情况下的交通诱导策略和用户最佳路径、出发时间选择 On
21、fully inter-connected, traffic guide strategy and user best route and departure time choice运输者：最大化地使用资源 operators: maximize the use of resources如公交、地铁、自行车，无缝接驳 Example: bus and subway, bicycle, Seamless connection客运和货运分时分段使用道路 Segmented using road passenger and freight,32,思考：对城市、交通规划的影响,交通与城市、社交紧密联动
22、 Interactive between transportation and urban, social居住地、工作地的选择 choice of residence, workplace工作方式、生活方式的选择 choice of work and live style社交活动的时间、地点安排 position and time arrangement of Social activities餐饮、旅游、购物的营销管理、总量控制 marketing management and amount control of Catering, tourism, shopping,医疗、教育、娱乐等设置
23、的部署和运营 deployment and operations of hospital, education, entertainment,33,Influencefor urban, transportation planning,思考：对城市、交通规划的影响,交通要素之间、与城市和社会之间相关关系量化 quantitative the relationship between transportation elements and urban, society交通系统涉及因素非常多，而且往往交错、叠加影响 many elements, and Staggered, superpositi
24、on interact with each other系统动态变化，上下游、前后时间段相互制约 dynamic changing, limited with Upstream and downstream, before and after实时、稳定的大数据计算和分析能力 Real-time, stable big data computing and analyze capacity交通关系民生，数据计算和处理质量要求高 Relation to livelihood, calculation and processing higher quality requirements以人（车）的活动为最基本的数据单元，数据规模庞大 Person (car) activities as the basic unit of data, data size is huge,34,挑战challenge,对交通需求内在规律的重新认识和建模对交通系统运行规律的重新认识和建模对交通流传播的重新认识和建模,35,大数据应用的核心,谢谢！,
