Study On Identification Of Key Sections Of Urban Road Network Based On Multi-source Information Processing Technology

With the rapid growth of urban car ownership year by year,the contradiction between supply and demand of road traffic has become increasingly prominent,which leads to more serious traffic congestion.The difference between the static attribute characteristics and traffic flow characteristics of traffic sections makes the role and importance of each section in the whole road network different.Once some key sections fail(such as traffic accidents,construction and traffic jams,etc.),it will reduce the connectivity of the road network,affect the overall performance of the road network,and even cause a large area of traffic congestion.The development of intelligent detection technology and its wide application in the field of transportation provide a variety of traffic data for the traffic system,and provide more comprehensive and rich information for the identification of key sections of urban road network.Therefore,combined with the multi-source information processing technology with fast response speed and strong self-learning ability,it can provide effective decision support for the optimization of road network structure,traffic management control,emergency evacuation and rescue by using the massive traffic data collected from multiple detection sources to identify the key sections of urban road network.Firstly,from the demand of static and dynamic data for key road section identification,this paper introduces the traffic data acquisition technology and its advantages and disadvantages,analyzes the basic characteristics of multi-source dynamic traffic data,introduces the calculation method of common parameter indexes according to the parameter selection principle of traffic flow basic theory,and expounds the application advantages and disadvantages of multi-source information processing technology in the identification of key road sections of urban road network The classical calculation models and their advantages and disadvantages are introduced.Secondly,according to the abnormal situation of GPS data of floating car,the method of data processing and map matching based on FME is proposed,and the temporal analysis of GPS trajectory data is carried out by using Arc GIS.on the basis of fault identification of fixed detection data by using threshold method combined with Mahalanobis distance,the fault data prediction and repair model based on GM(1,N)-GA-WNN is constructed.In view of the limitation of single source data and the complementarity of multi-source information,a multi-source data fusion model based on MEA-BP is designed under the premise of space-time matching,combining the global search of mind evolutionary algorithm and the adaptive learning advantages of BP neural network.The example shows that the repair and fusion model can get high quality and high coverage basic data of key road section identification at a fast speed.Finally,considering the propagation and directionality of traffic flow,the original urban road network model based on directed weighted complex network is constructed.according to the structure characteristics of road network and traffic parameter data,a reasonable evaluation index system of key road sections is constructed based on quantifying the interaction degree of traffic flow between road sections and adjacent road sections combined with spatio-temporal correlation coefficient,and the main road sections are comprehensively considered objective factors optimize the distribution of index weight,and propose a key section identification model of urban road network based on AHP-Entropy-TOPSIS.The propagation dynamics model of variation SI is used to verify the feasibility of the key section identification method.The case study shows that the model based on multi-source information processing technology has strong stability and high accuracy.