Research On Reliable Travel Mode Identification Based On Mobile Phone Signaling Data

The research on the reliable identification of traffic travel modes is an important basis for traffic planning based on the "four-stage method",which has a positive effect on urban traffic planning and control,optimizing the structure of traffic travel modes,and alleviating road traffic congestion.However,in the face of increasingly abundant transportation modes,existing travel mode identification research institutes use traditional manual questionnaire data and mobile phone GPS positioning data,which have problems such as high data acquisition costs and limited data samples.The development of mobile communication technology provides a lowcost,large-sample data source for the research on traffic travel mode identification.Existing researches on using mobile phone signaling data to identify travel modes mostly use deterministic models such as rule-based models,machine learning models,and statistical analysis models.There are problems such as insufficient consideration of the uncertainty of traffic travel characteristics and insufficient utilization of incomplete mobile phone signaling data.Based on this,based on the mobile phone signaling data,this paper deeply analyzes the time specificity and uncertainty of traffic travel characteristics,and adopts the Bayesian network model to solve the problem of reliable identification of traffic travel modes based on mobile phone signaling data.Three achievements have been made: quantitative characterization of characteristic time specificity,construction of a reliable identification model of travel mode considering the uncertainty of traffic travel characteristics,and travel mode identification for incomplete mobile phone signaling data.The main work of the paper is as follows.Traffic travel feature extraction and time-specific analysis.Effective traffic travel feature extraction and travel feature time-specific characterization and analysis are important prerequisites for reliable identification of travel modes.Existing travel mode identification researches extract travel feature categories from mobile phone signaling data limited to travel behavior characteristics,and Time-specific analyses of travel characteristics are insufficient.Based on these two deficiencies,on the one hand,this paper extracts multi-dimensional traffic travel characteristics from three perspectives: traffic environment attributes,traveler personal attributes and traffic travel behavior attributes based on the location of mobile phone base stations and mobile phone users’ signaling data.Based on the distribution characteristics of duration characteristics and the uncertainty of travel speed characteristics distribution,an index to quantify the time specificity of traffic travel characteristics is constructed,and the time period distribution characteristics of various travel modes are clarified,which lays the foundation for the construction of subsequent travel mode identification models.Construction of a travel mode identification model considering the uncertainty of traffic travel characteristics.The different operating states of the transportation system and the different physical properties of the vehicles make the travel characteristics of various travel modes show many uncertainties.Based on the conclusion of the dependence relationship between the traffic travel characteristics in the sense,a reliable travel mode feature causal relationship network is constructed,and the travel mode identification model based on the Bayesian network model is completed.The overall identification accuracy of the model is ideal,and the grasp of the urban traffic travel mode structure is relatively accurate,which can provide a basis for traffic planning and management,and traffic travel mode structure optimization.Robustness analysis of travel mode identification models.The level of mobile communication hardware facilities and the instability of the communication process make the use of mobile phone signaling data to conduct traffic research with certain uncertainties.The specific manifestation is that there are different degrees of data missing in the mobile phone signaling data.Therefore,it is necessary to evaluate data scenarios with different completeness.The performance of the following model is of great significance for the reliable identification of travel modes.In this part,based on the possible lack of mobile phone signaling data in reality,three observation data scenarios with different degrees of completeness are constructed by combining multi-dimensional traffic travel characteristics,which verifies the robustness of the constructed travel mode identification model,and focuses on analyzing The influence of three traffic travel behavior characteristics on the correct rate of travel mode identification,the key travel characteristics in the process of travel mode identification were mined,and the relevant conclusions pointed out the direction for reliable identification of travel mode using incomplete observation data.The research results of this study can help transportation planning and management departments to more accurately grasp the structural characteristics of urban transportation modes,and provide an important basis for developing green transportation,low-carbon transportation,and formulating optimization policies for transportation mode structure.