Study On The Multi-mode Travel Integration Method Of City Group Driven By Spatial-temporal Trajectory Data

With the wide application of Global Navigation Satellite System(GNSS)and Geographical Information System(GIS)technologies,the acquisition of large-scale individual spatiotemporal data is becoming a reality.The use of traffic trajectory big data to extract city groups' travel status,job relations,and travel patterns has become a hot topic at home and abroad.Urban community trips and the spatial interactions they contain are the basis for the analysis of human activities and the study of urban spatial structure.Under the influence of globalization and informatization,trajectory big data and artificial intelligence methods combine to bring new approaches for urban research and human activities analysis.Urban traffic includes multiple travel modes.A single travel mode only represents the pattern of travel of some groups in the city.It is biased and cannot represent the status of the entire city.The data generated by different models shows different patterns of travel and other information.The law of urban group travel plays an important role in urban transportation planning and management,land use layout,urban infrastructure,and urban traffic forecasting.However,the study found that most scholars currently use single travel mode data to analyze the patterns of urban group travel.Integration of easily accessible travel laws to get the overall travel information of urban residents has brought new research content to geography research.In this paper,we use the bus,metro,taxi,and mobile phone to explore the multi-modal travel integration method.The multi-modal travel integration in this paper is not a simple combination of multiple types of travel data,but a fusion of travel law of multi-mode travel.Taking the travel pattern as the characteristics of the travel data,the fusion method in the paper can refer to the method of feature level fusion of sensor data.The main research idea of this paper is to use multi-mode travel data,based on the analysis of travel correlation and travel density,and to propose a multi-modal travel law fusion framework based on sensor data feature level fusion method to obtain the fusion coefficient of multi-modal travel fusion.This article uses the 700 million taxi GPS,bus card GPS and smart card track records in Shenzhen every week to extract travel records of the city's taxis(Taxi,T),buses(Bus,B)and metro(Metro,M),and use travel volume.,distance,time and direction as travel indicators to carry out multi-model travel correlation test.The experimental results show that:(1)The consistencies of bus-taxi,metro-taxi are quite low while the consistency of bus-metro is significantly high.(2)There are spatial heterogeneity and anisotropy between the consistent multi-mode urban travels.The consistency of taxi,bus and metro travels are high in the urban center,but low in the suburbs.Also there is a significant anisotropy behind the high consistent travels.All three travels are highly correlated in the west-east direction and the south-north direction.Based on the results of the above-mentioned test for the significance of correlations,and combined with the travel density,Shenzhen is divided into 5 different large regions.The travel records obtained from the mobile phone data are used as the overall travel records of Shenzhen City,combined with the method of indirect adjustment.The fusion coefficients of travel modes in different regions were obtained and compared with the urban traffic structure.The experimental results show that this fusion method is feasible in most areas of Shenzhen.Under the influence of globalization and informatization,trajectory big data and artificial intelligence methods combine to bring new approaches for urban research and human activities analysis.Multi-modal travel integration improves the week of insufficient data in urban group travel research,providing data support for in-depth study of urban group travel and urban transportation.Also it lays the foundation for future urban construction,urban planning,urban land use and even urban development,and provides suggestions for further Shenzhen's transport facilities improving and solving traffic congestion problems,and methods for data fusion and promote scientific development.