Semantic Road Network Construction And Path Planning Oriented To GPS Trajectory

There are a lot of advantages of GPS trajectory data,including wide coverage,fast update,easy collection,low cost,especially implies rich information of road network.The user's GPS trajectory data has gradually become a hot research topic for scholars all over the world as an emerging data source for constructing urban road network.The thesis focuses on the key issues of GPS trajectory data processing,such as build the city road network topology based on GPS trajectory data,model the Semantic Road Network,plan the path for users based on semantic network.The main research contents of this thesis are as follows:1.GPS trajectory data preprocessing.The preprocessing includes two aspects,"Trajectory filtering" and "Trajectory segmentation".As for the "trajectory filtering",this thesis applies the mean filtering algorithm based on the velocity,to denoise and smooth the obviously deviated GPS trajectory points;and this thesis divides the GPS trajectory into "trip segments" and "activity segments" according to the velocity and angle change rate,in order to indicate the use's trajectory along the road network and in the region respectively.2.Propose a new identification method of intersections based on GPS trajectory.The intersection plays a very important role in describing the connectivity of the road.Aiming at the existing problems in the identification of intersections,including single traffic mode,difficult threshold setting and diversification of intersections,in this thesis,a new intersections identification method,which combines the Geohash algorithm and the KNN algorithm,is proposed,in order to identify "trip segments".The first step is to uniformly encode user's trajectory data and activity area with the Geohash algorithm,then to select and label the intersection feature training set,and finally to identify the intersections by the classification idea of the KNN algorithm.This method is not affected by the traffic mode,sampling frequency change and urban road form,and the experiment on real trajectory data set shows that the intersection identification method proposed in this thesis has high recognition precision and strong versatility.3.This thesis presents a method based on GPS trajectory for semantic road network construction and path planning.The construction of road network is the foundation of pathplanning,and the construction of semantic road network adds more semantic features to the road network.Firstly,on the basis of "road network intersection identification",this method carries on the serialization processing and the connectivity analysis to the GPS track,constructs the road network topology;then enriches the road network topology semantic weights by deeply excavating the semantic attribute of the topological section and the number of POI types in the section area.Finally,on the basis of semantic road network,make path planning by Cons_dijkstra algorithm,to provide more optimized path selections for users.4.In this thesis,the prototype of visual analysis system based on GPS trajectory is designed and implemented.Combined with the algorithm in this thesis,the system is able to intuitively express the information on the map of GPS trajectory.The experimental results show that the system can be an important auxiliary means to study GPS trajectory,and has strong practicability.In summary,the method for intersection identification,the semantic road network construction and the path planning based on GPS trajectory presented in this thesis,combines the geographic information technology with the computer technology,and solves the problem of constructing semantic road network based on GPS trajectory.This research has wide practical significance and broad application prospects.