Map Matching Algorithm And Application Research Based On GPS Vehicle Trajectory Data

With the development of city,traffic congestion situation is more and more serious,which has affected people's lives and has caused a huge waste of resources.In order to relieve traffic congestion,most countries have studied Intelligent Transportation Systems(ITS).Map matching is an important part of the intelligent traffic system,and the existi'ng map matching algorithms cannot guarantee a high matching accuracy in dealing with low-sampling-rate GPS vehicle trajectory data(the sampling interval is greater than 1 minute).To solve this problem,this paper proposes a map matching algorithm for low-sampling-rate vehicle trajectory data,and applies the matched vehicle data for traffic congestion analysis.The main work and contributions of this article are as follows:(1)Firstly,this paper systematically introduces the research background and significance of map matching algorithm and traffic congestion,and summarizes the research status and characteristics at home and abroad.Secondly,the basic principle of map matching algorithm is emphatically analyzed,and compares the existing map matching algorithms.Then,the map matching preprocessing technology is introduced.In addition,the related concepts of traffic jam and the temporal and spatial characteristics of traffic jam are introduced.(2)For the traditional map matching algorithm dealing with low-frequency vehicle trajectory data,it can not guarantee the higher matching accuracy.This paper proposes a directed map matching algorithm based on hidden Markov model.Traditional algorithms based on Hidden Markov Model only consider the positioning error of the observation point when calculating the observation probability.When the initial moment of observation point is at the intersection,it will not be able to match the exact road segment.Therefore,when calculating the observed probability,this paper considers the driving direction of the vehicle.In addition,this paper introduces the geometric characteristics of the road network and the change trend of the positioning error into the observation probability model to further improve the accuracy of map matching.Experimental results show that the proposed algorithm can maintain high matching accuracy when dealing with low-sampling-rate and high-sampling-rate vehicle trajectory data.(3)This paper applies the matching vehicle data to analyze the spatio-temporal characteristics of traffic jams.According to the characteristics of vehicle trajectory data and traffic congestion theory,the average speed of the road segment is calculated and traffic congestion is classified,and the traffic congestion changes in different periods and different regions are analyzed.The analysis results show that the regularity of traffic congestion is strong and often occurs in the urban center.