大数据与城市空间活动分析.pptx

大数据与城市空间活动分析提要与传统理论的衔接与变革构建基本数据模型嵌套式的证析框架城市空间联系结构分析与传统理论的衔接与变革Bottleneck of technical means: the collection of continuous dataThe traditional transportation survey carried out every 5-10 years fails to meet the needs of process management of urban transport policy. As for urban transportation, big data is not vogue but value, embodied with technology push, demand pull and idea revolution.Data Environment and Decision SupportThe recognition of the current situationThe understanding of the laws of historical developmentThe analysis on the system patternsClarify the structure of problems and define the objectives of strategiesPlanningEvaluationRolling correction in the implementationTravel behavior observation analysisTraffic state observation analysisDATAFCDBus operationRail-transit operationRoad detectionLicense plate recognitionDATAIC cardTaxi operationOperation of parking lotsCellphone dataLicense plate recognitionMulti-source data fusionMulti-source data fusionTravel behavior patterns extraction and analysisTraffic state patterns extraction and analysis与传统理论的衔接与变革传统交通调查数据“今天大多数人去哪里”，连续追踪数据表征的是个体“经常去哪里”；并非定制数据，难以准确区分“一次出行”，但是可以利用时间活动积累，说明经常活动的地区；对于出行目的等只能够通过活动规律估计。移动通信数据与区域出行总量的关系移动通信数据?区域出行总量：可以利用较长时间段内的移动通信数据，计算日平均活动点个数，对区域交通出行需求进行估算日平均活动点个数?出行率：活动点可以反映传统的居民出行调查中遗漏的一部分附带活动研究区域1陆家嘴街道2南京东路街道3五角场街道4真如镇5徐家汇街道6吴淞街道7嘉定镇街道8顾村大居9江桥大居10泗泾大居移动通信活动点与居民出行D点空间分布一致性移动通信数据与调查数据Kolmogorov-Smirnov检验统计量侧检验显著性概率p0.05，证实了两样本数据具有相同分布的假设。移动通信数据与传统交通调查数据在出行次数、D点分布等方面具有可转换性，可以利用这一特点补充传统交通调查，进行两次调查时间之间的数据内插。Analysis based on new data environment has connection with traditional frameworkActivity points (mobile phone data)? Trip rates (comprehensive transport survey)Analysis of spatial interaction based on frequent item setTrip rates(comprehensive transport survey)For the connection with traditional experience, relationship should be established between the new method and the traditional framework. To make full use of the pervasively monitoring data, innovative approaches for feature expr