Research On Generation Of Travel Chain And Extraction Of Travel Feature Based On Smart Phone

With the rapid development of the economy and the acceleration of the urbanization process,the number of urban population and motor vehicles is increasing,and the problem of traffic congestion in cities is also becoming more and more serious.It has become the consensus of more and more scholars to carry out traffic survey of urban residents to obtain relevant traffic data,and then analyze and study these data.The results of data analysis are used to carry out urban traffic construction and planning,and then solve urban congestion problem.The passive GPS survey method is used more and more widely in travel surveys because of its high positioning accuracy,low burden on volunteers,and ability to collect data continuously.GPS data collected through travel surveys are only a series of time-fixed positioning points.Simply observing these positioning points does not provide information related to traffic travel.Through the analysis and processing of GPS data by certain methods,the relevant information of the traffic travel hidden therein is discovered,and the status of urban traffic and the characteristics and needs of residents' travel are obtained,so as to provide data support for urban traffic planning and construction.Therefore,the information mining and processing of GPS trajectory data is a meaningful and important task.In order to intuitively understand the hidden information in the trajectory and mining trajectory,this report has conducted some research on the analysis and processing of trajectory data.The main contents include:(1)This report describes the smart phone-based travel trajectory acquisition and denoising methods.According to the demand of trajectory analysis,this report constructs a sampling format for storing travel trajectory data,and designs a kind of denoising method according to the distance,speed and time threshold according to the different characteristics of the deviation of GPS data collected by mobile phones.(2)For travel trajectory data collected by smart phones,the report proposes a travel stay point recognition algorithm,based on which a travel trajectory chain is generated.In this paper,the definitions of travel chain and stay point are given.The types and characteristics of stay points are analyzed.A real-time stop point recognition algorithm based on DBSCAN(Density-Based Spatial Clustering of Applications with Noise)is proposed.The algorithm can identify indoor stays and outdoor stays,set time thresholds to reduce the number of clustering,and then use the actual data to verify the effectiveness and accuracy of the algorithm.At last,the report builds a travel chain based on time series,travel track points,and stay points.(3)For the problem of missing trajectory in travel data,this report proves that it is NP-Hard problem.Based on graph theory and edge coloring mechanism,we propose an improved Dijkstra trajectory completion algorithm.The algorithm abstracts the road network into a weighted undirected graph,then colors the edges according to different travel modes,and uses the timing constraints of the source and the end points in the graph,which greatly reduces the range of feasible paths for searching and speeds up the operation of the algorithm.After that,the trajectory points are acquired from the selected path,and parameters such as time and positioning type are assigned to them,and finally saved to the database to form a complete trajectory data.(4)Combined with the collected real GPS trajectory data,this report proposes a trajectory data feature analysis and extraction method.Based on the identified stop point,this report studies the method of identifying the types of residential building for travel destination,and extracts the personal characteristics of travellers and travel characteristics of GPS trajectory data.