Dynamic Detection And Update Algorithm For Changes Of Road Network Based On The Traffic Trajectory Data

How to detect the rapid changes of road network in time is one of the major issues for LBS services.A large number of real-time traffic trajectories often coincide with the latest traffic network,which includes the dynamic information of urban road network.Nowadays,it has become a hot research topic that detects the changes of the road network using the data of floating car trajectory data.Due to the frequently development of road infrastructure in Pingtan Island and the increasing optimization of urban road network structure in Fuzhou city,a new dynamic detecting method is proposed in this paper for the road network updating based on the traffic trajectory data.The main work is as follows:1)The research status is investigated about the traffic trajectory data mining in the field of road network updating,and the key problems are discussed including data preprocessing,map matching and accuracy assessment of change detection in road network change detection process,and then the system implementation framework is designed finally.2)An optimized algorithm is proposed for map matching,and an adaptive matching method of grid-segment inverted index is constructed,and then a weighting model and a matching algorithm of road network topological relations are put forward for the detection object of road network,which realizes the accurate data of traffic track data and provides a good basis for road network change detection based on traffic trajectory data.3)A new road detection method is presented which integrates track data of engineering vehicle and taxi.Firstly,new roads and other building sites are detected through the engineering vehicle trajectory,then the candidate new roads are verified through the taxi track,which means it has been put into use,and some concept related to clustering are introduced,so as to achieve rapid and accurate new roads detection.4)Based on the improved map matching algorithm,this paper proposes a trajectory feature analysis method with vehicle steering angle.At the same time,a mathematic model is constructed for the geospatial data.Based on the existingnetwork topology information,detection and identification of the change of the intersection steering rules is realized.Finally,combined with ArcGIS and other software platforms,the above methods were developed and validated.Experimental results show that the algorithm proposed in this paper can quickly and timely detect changes in road network information to meet the requirements of traffic information service time and efficiency.