Research On Identification Of Critical Nodes In Urban Traffic Network Based On Mobile Phone Signaling&Its Spatio-temporal Characteristics

With the continuous improvement of urbanization level,the scale of urban traffic network is expanding,while,the disaster risk faced by the traffic network is increasing too.Facing all kinds of disasters,the vulnerable defense ability of traffic network is constantly emerging,"cascade failure,paralysis and collapse" phenomenon often occurs,causing immeasurable losses to the city.Therefore,there are great challenges for traffic managers to achieve controllable and strong adaptability under the existing resources.The development of mobile communication technology has brought the application era of mobile phone signaling big data,and produced massive and high-precision resident location informations and trajectories,which has provided possibilities for the comprehensive perception of urban traffic network.Taken the traffic network intersection node as the research object,the management and protection of the road intersection node as the breakthrough point,used the massive positioning data generated by mobile signaling to identify the critical nodes of the urban road traffic network.Started with large sample residents travel activities,the spatial and temporal characteristics of residents’ activities on critical nodes were analyzed.Some scientific and reasonable critical nodes managements and unblocked strategies were formulated based on these,which can enhance the reliability of the urban traffic network and enhance the network response ability.Specifically,the main contents of this thesis are as follows:1.To solve the obstacle of mobile signaling in traffic data acquisition application,the basic program of extracting vehicle trajectory data on urban road network based on mobile signaling data was established.An 3D ray propagation model based on KNN of road precise matching algorithm and different trajectories pattern recognition method using time interval,sequence length and speed as clustering thresholds were proposed.To eliminate abnormal trajectories and ensure the quality of data,it had put forward to "speed,distance constraints,time constraints," three indicators as evidence identifying abnormal tracks detection method of evidence theory.Thus,accurate and reliable spatio-temporal trajectory data of vehicle-mounted users on changsha urban road traffic network were obtained.2.Due to the diversity of identifying methods of critical nodes in complex network,throuh the comparative analysis of existing methods.the definition of key nodes in urban traffic network with "five features" based on mobile signaling trajectory data,and the"six-step method" to identify critical nodes in urban traffic network were proposed.The reliability evaluation method of information fusion is established for the results of critical nodes.To analyze the spatial and temporal characteristics of critical nodes in depth,it had broken the limitation of traditional framework urban traffic analysis,established a multi-dimensional description system of residential activities in critical nodes based on"residential activity space analysis",put forward 11 indicators for quantitative description of spatial and temporal characteristics,and holographic described the mechanism of residential activities,spatial and temporal evolution and spatial on critical nodes.The spatial and temporal characteristics of the "traffic source" of the traffic network node were profoundly revealed3.To verify the feasibility of the theoretical method of obtaining critical nodes of urban traffic network based on mobile signaling data.In Changsha city as an example,based on Changsha mobile signal data,combined with the definition and critical node identification method,the critical nodes(72)of the urban road traffic network nodes(851)in Changsha City were successfully identified and extracted.and the reliability of critical node identification results is determined through the reliability discrimination system constructed by "four big data sources".Six typical critical nodes were selected for the temporal and spatial characteristics,and the description and analysis of the temporal and spatial characteristics of the critical nodes were carried out from eight dimensions.Finally,11 targeted and unblocked countermeasures for critical nodes of Changsha traffic network were put forwarded from the macroscopic,mesoscopic and microscopic perspectives,to achieve effective management and strong protection of Changsha traffic network nodes.