Research And Application Of Regional Traffic Pattern Recognition Based On Mobile Signaling Data

Urban agglomeration traffic planning and construction is an important part of seeking regional integration and optimizing the development of urban agglomeration.In order to meet the practical needs of urban agglomeration traffic planning,the extraction of regional traffic feature information and the construction of regional traffic model have always been the focus of research.However,due to the difficulty in obtaining data resources of urban agglomeration scale,relevant researches mostly remain at the theoretical level and fail to make substantial progress.With the development of mobile phone big data technology,using mobile phone location technology to extract regional traffic characteristic information breaks the limitation of traditional resident travel survey,and brings opportunities to break through the bottleneck of regional traffic research.However,the research on regional travel pattern recognition and application based on mobile signaling data is not enough.In this context,this paper puts forward the identification model of residents’ regional travel mode according to the characteristics of regional travel road conditions and communication environment.The model can effectively identify the three regional travel modes of high-speed rail,car and coach and the travel paths of residents.After that,this paper designs a regional travel experiment to demonstrate the reliability of the algorithm model recognition results,and uses this model to carry out an example analysis of regional travel characteristics in Chenghua district of Chengdu city and Yanjiang district of Ziyang City.Finally,this paper applies the sharing ratio data obtained from the mode recognition model to load into the time value model based on mode choice,combining with travel time data,travel cost and other data,and explores the relationship between the time value of regional travel residents in Chenghua district and Yanjiang district and their choice of travel mode.Firstly,this paper introduces the mobile communication technology and mobile phone positioning technology.This paper defines the boundary of the region’s external travel,and introduces the path matching technology and data sample expansion technology of mobile phone signaling data.Secondly,this paper designs a regional travel experiment to collect real travel signaling data.The data characteristics of regional travel cell phone signaling data,distribution characteristics of regional travel connection signal base stations and travel speed characteristics are analyzed,which lays a foundation for the construction of regional travel mode identification model in the following paper.Then,a regional travel pattern recognition model is constructed by integrating trip path information,trip average speed,trip endpoint and other information.The model first proposed a travel path recognition model based on Needleman-Wunsch algorithm to extract residents’ travel path information,and initially divided residents’ trips into road trips and high-speed rail trips by using the travel path information.Select the residents whose travel path is the same as the coach route from the residents.The average travel speed and the latitude and longitude information of the travel endpoints were extracted by the spatio-temporal clustering algorithm.Finally,the rule-based fuzzy identification model of road travel mode was constructed by integrating the above extracted residents’ travel characteristics,and the road travel was further divided into car travel and coach travel.After that,this paper analyzes the reliability of travel path and travel mode identification algorithm by using the collected mobile phone signaling data samples,and then applies the algorithm to carry out macro analysis and verification of travel characteristics in Chenghua district of Chengdu and Yanjiang district of Ziyang.Finally,this paper introduces the meaning,classification and characteristics of travel time value.Combined with mobile phone signaling data to identify travel mode sharing ratio data,travel time data,travel cost data and other data to calculate travel time value,the relationship between travel time value and travel mode choice of residents in Chenghua district of Chengdu and Yanjiang district of Ziyang City was studied.