Research On Travel Characteristics Of Urban Residents Based On 4G Mobile Phone Signaling Data

With the development of intelligent collection technology,the research on the travel characteristics of urban residents based on multi-source data has gradually become the mainstream.Mobile phone signaling data has become a research hotspot to promote the development of smart cities because of its huge data volume and wide coverage.Studying the travel characteristics of urban residents based on mobile phone signaling data makes up the shortcomings of traditional survey methods such as high collection cost and limited sample collection.Considering that the current communication base stations are mainly 4G communication base stations,this paper studies the travel characteristics of urban residents relied on 4G mobile phone signaling data,mainly from the following three parts:Firstly,this paper proposes a data preprocessing method the massive mobile phone signaling data and constructs a stop point identification algorithm based on the DBSCAN clustering algorithm extracting the effective travel trajectory data of urban residents.Considering the generation principle of abnormal mobile phone data,combined with the positioning characteristics of the base station,it can complete the data merging and deletion preprocessing and eliminate the noise data generated in the process of residents’ travel.After limiting the parameter range of DBSCAN algorithm on the basis of the positioning accuracy of4 G base station,this research iteratively calculates and evaluates the clustering effect under each parameter combination and finally determine that the distance threshold parameter of DBSCAN clustering algorithm is 400 m,the minimum number of points in the class is 5.On account of the clusters of residents’ stay points,the stay center points are identified with the principle of minimizing the mean distance from the point to other points in the cluster.Taking the unclassified move points into consideration,the residens’ effective travel trajectory points are gathered in chronological order,which provides a data basis map matching of resident travel data.Secondly,this paper proposes a hybrid map matching algorithm based on mobile phone signaling data and selects the search radius and the maximum number of candidate points by comparing and evaluating the operating results of various map matching algorithms.After,this paper analyzes the advantages of the map matching algorithm proposed.Combined with the Dijkstra shortest path algorithm,the complete travel paths of residents can be collected.In view of the principle of map matching based on positioning data,integrated with the idea of geometric matching and topology matching,a hybrid map matching algorithm related to Hidden Markov Model has been constructed and the trajectory point sequence of residents’ travel can be extracted according to the Viterbi algorithm.The search radius and the parameter range of the maximum number of candidate points are selected,through the algorithm accuracy and running time,by comprehensively comparing and evaluating the map matching effects of the ST map matching algorithm,the intersection-based map matching algorithm and the algorithm proposed in this paper,which are carried on the same verification data under different parameter combinations.and determine that the search radius is 250 m,and the number of candidate points is 4.Under these parameters,the permance of the map matching algorithm proposed in this paper is basically better than the other two algorithms,with an accuracy of more than 90%.Combining the Dijkstra shortest path algorithm and the map matching algorithm proposed in this paper,it is feasible to extract the complete travel routes of residents in Kunshan and provide actual route inmation the analysis of residents’ travel characteristics.Finally,this paper proposes a method extracting the spatio-temporal distribution characteristics of residents’ travel based on the travel paths of residents with spatio-temporal inmation.After,this paper constructs a quantitative model of travel characteristics in urban residential areas and analyzes the travel characteristics of residential areas relied on road networks.According to the characteristics of staying points and moving points,this paper extracts the basic characteristics of residents’ travel,such as travel hotspots,travel distance,and travel times.A road segment-based travel congestion feature classification model residential areas was established,and the road segment travel congestion feature index was quantified to further analyze the travel characteristics of regional residents on the road network.