Research On Optimal Path Planning Based On Taxi Trajectory Data Mining

With the wide use of GPS navigation and intelligent positioning terminal,it is possible to obtain high precision trajectory data.At present,most of the city's taxis have been loaded with GPS equipment to collect the GPS track data generated by the taxi.The path planning model based on the trajectory data generated by taxi can get the optimal path between any two points in the city,which has important guiding significance for the travel of the vehicle.Because of the traditional path planning method does not take into account the experience knowledge in taxi trajectory data,on the basis of urban road network information,the travel time and path of the vehicle were calculated,and regarding the shortest path of time and distance as the shortest path.Therefore,how to use a large number of GPS trajectory data for data mining to achieve the optimal path planning is the key problem to be solved.Aiming at the above problems,a path planning method based on hot spot road map and vehicle travel time is presented.In this method,the path planning technology,which is based on the computation,has been changed to data driven data mining technology,therefore,it is possible to obtain more practical optimal path.The main contents of this paper include:(1)Preprocessing of map data and trajectory data.Aimming at the data acquisition of the electronic map,the Open Street Map(OSM)platform can be used to obtain the road network data.Aimming at the acquisition of taxi trajectory data,there is an error in the information acquired by the GPS positioning device,so it is necessary to filter and extract the trajectory data.On this basis,aimming at the low sampling rate feature of GPS trajectory data,An Interactive Voting-based Map Matching Algorithm(IVMM)is adopted,which is used to realize the map matching of trajectory data to road network data.The experimental results show that the algorithm can effectively improve the accuracy of trajectory data matching to network data.(2)Construction of hot section map.Aiming at the establishment of the road network layer,hot section map with the experience of taxi driving and the driving time of the road is constructed.First,according to the track data to the road network data map matching results,select nodes of the hot section map.Secondly,in order to realize the effective connection between different nodes,an improved trajectory clustering algorithm based on editing distance is proposed.Then,the time characteristics of the trajectories between nodes are clustered and analyzed.Finally,Finally,we give the time weight for the connection between different nodes,and establish a complete hot section.(3)A path planning method based on hot section map.Aiming at the research of path planning method,an improved path planning method based on hot section map is proposed,and compared with the shortest path planning method based on basic road network and the path planning method based on hierarchical road network.The experimental results show that the path obtained by the path planning method based on the hot section map is more in line with the optimal path in the actual travel time.The main contribution of this paper is to fully excavate the empirical information in the trajectory data,to construct the actual driving trajectory as the basis of the empirical road network construction,and propose a new network layer structure to provide effective support for the path planning.