Travel Characteristics And Travel Mode Identification Based On Mobile Signaling Data

In recent years,my country’s per capita car ownership has increased year by year,and domestic traffic congestion has occurred from time to time.Related departments are constantly seeking solutions to improve the bad traffic phenomenon.Mobile phone signaling data is the data that reflects user information generated when signaling is transmitted between different links of the mobile communication network.After desensitization and sample expansion,it can be applied to residents’ behavior preferences and urban traffic planning layout.Research in other aspects is one of the main data sources for research in the field of intelligent transportation at this stage.Based on the mobile phone signaling data provided by operators,this paper analyzes the travel behavior of residents in Changchun City from the needs of urban traffic management planning.The specific research contents are summarized as follows:1.Extract user travel trajectory.Aiming at the problem of the positioning deviation of the base station’s positioning coordinates,after performing operations such as deduplication,error removal,drift removal,and ping-pong data on the mobile phone signaling data,the point-to-point matching method in the geometric analysis method is used to match the base station’s positioning coordinates.The spatio-temporal clustering algorithm recognizes the staying and moving state of the user’s trajectory points,and considers the two staying points and multiple moving points between the staying points as the moving trajectory of the user in a single trip.2.Analysis of residents’ travel characteristics.Based on the user movement trajectory extracted in the previous part,the characteristics of residents’ travel time and space and travel intensity are analyzed.First,the user’s hourly travel volume is calculated to show the fluctuation of the user travel volume with the travel time,the global and local Moran index is calculated,and the mobile terminal is observed.The overall law of users in mobile gathering behavior.Then count the number of residents trips under different travel distances,count the average number of trips of users,calculate the average travel time of residents that day,and analyze the characteristics of residents’ travel intensity from three aspects: travel distance,travel frequency,and travel time.3.Identification of residents’ travel mode.The travel characteristics are extracted from the user’s travel trajectory,and the initial membership function is constructed based on the prior knowledge of the three characteristics of travel distance,travel time,and travel speed,and considering that the Mahalanobis distance is more sensitive to the changing relationship between multi-dimensional data,This paper compares the traditional Euclidean distance with the Mahalanobis distance,and proposes a resident travel mode recognition model based on the improved fuzzy C-means clustering algorithm.The three common travel modes for users in the urban area of Changchun,namely walking and bicycle,Motor vehicles are identified.Based on the extraction of travel trajectories from mobile phone signaling data,this paper analyzes the travel behavior of residents in Changchun from two aspects: travel feature analysis and travel mode identification.It aims to improve the current complex traffic environment and has profound practical significance.