Traffic Flow Characteristics Analysis And Congestion Identification Of Holiday Highways Based On Multi-source Data Fusion

As an important part of road transportation,the ETC(Electronic Toll Collection)system,which allows non-stop electronic toll collection on highways,has matured in recent years.At the same time,the free-flow toll system composed of gantries has been installed and put into use.Meanwhile,data-driven intelligent transportation based on big data has gradually become mainstream,and data is the foundation for the development of intelligent transportation.The toll stations and gantry systems can record the driving information of passing vehicles,providing a data basis for understanding the operating rules of highways during major holidays.In this article,based on toll data from toll stations,toll data from gantry systems,and data on passengers and goods transportation,a multi-source data trustworthy fusion method based on evidence theory is proposed,combined with fuzzy theory.The credibility of single-source data and the support of multi-source data are calculated to fuse the travel time extracted from the three types of data,providing reliable road segment travel time for subsequent theoretical calculation and analysis.Next,using existing data,traffic on highways is analyzed based on the time and spatial distribution of traffic volume during holidays and non-holidays,including the annual average,monthly average,and daily average distribution of traffic volume over time,as well as the flow distribution,destination distribution,and distance distribution of traffic volume spatially.By comparing and analyzing the traffic volume distribution during holidays,and understanding the mechanism of holiday congestion,a theoretical basis is provided for subsequent traffic congestion identification and road operating status evaluation.Finally,an evaluation index system for highway operation is determined,and the fuzzy Cmeans clustering algorithm is selected.By improving the algorithm and using the fuzzy algorithm of weight difference to determine an initial clustering center,the final clustering result is fixed.The weight of the clustering index is calculated using the feature weighting method,the analytic hierarchy process,and the entropy synthesis method,resulting in a stable and reliable clustering result.By determining the study area and dividing the road sections,the improved FCM algorithm can effectively evaluate the operating status of the highway network and display congested sections in the road network.