Research On The Evaluation Method Of Urban Road Service Level Based On Multi-source Data Fusion

In recent years,the number of cars in China has increased rapidly,which makes the problem of urban traffic congestion increasingly serious.Therefore,the full utilization of the existing road network resources and the improvement of road traffic control strategy are effective ways to deal with the problem of urban traffic congestion.How to make a reasonable assessment of urban road service level,so as to provide management basis for managers,is the basis of improving road traffic control methods.In addition,with the development of intelligent transportation system,urban roads are equipped with a variety of detection equipment,which can obtain different types and multiple dimensions of urban road traffic data,which provides sufficient data guarantee for the research of urban road waiter level evaluation.How to make rational use of the detection data of different detectors has become a very important part of urban road traffic management.Firstly,the fusion method of multi-source traffic data is studied.By analyzing the characteristics of different detection methods,the fusion level of multi-source data is determined as the fusion of feature level.Combined with the characteristics and correlations of various traffic parameters,the spatial characteristics,time characteristics and reliability of the data obtained by different detectors are analyzed,and the available traffic parameters are extended.On the basis of analyzing the spatial and temporal correlation of urban road traffic data,a variety of methods to complete the missing data are proposed to make the multi-source traffic data more complete.Then this paper studies the establishment of the index system of urban road service level evaluation.By analyzing the existing evaluation methods of road service level and the design methods of evaluation index system,it is determined to use tower structure for evaluation.The evaluation of urban road service level is divided into two parts: intersection service level evaluation and section service level evaluation.At the same time,according to the characteristics of all kinds of traffic parameters,these parameters are divided into intersection evaluation index part and section evaluation index part,which lays a foundation for the study of evaluation method in the following paper.Next,this paper studies the evaluation method of service level of urban signalized intersection.The multi-index evaluation method is determined to be used.BP neural network structure was used to establish the evaluation model,and the congestion index was obtained by using intersection saturation,queue length and average delay as indicators.Besides,the intersection equilibrium index model is established,and the evaluation grade of intersection service level is obtained by integrating the congestion index and the equilibrium index.Finally,this paper explores the evaluation method of urban road section service level.Firstly,the traffic flow characteristics of urban road sections are analyzed,and occupancy,flow and speed are selected as the characteristic parameters for the evaluation of road section service level.Meanwhile,the relationship among occupancy,flow and speed is analyzed.Then the SVM algorithm is introduced to train and verify the SVM model for traffic data of Wafangdian city,which indicates that the SVM model can effectively classify and evaluate the service level of urban road sections.