Research On Commuters’ Travel Mode Choice Behavior Based On Multi-source Data

The rapid development of social economy has promoted the process of urbanization,and the travel modes have come increasingly diverse,thus stimulating more travel demand and leading to increasingly prominent road congestion.Commuters are the main population of urban travel.Understanding the mechanism of their choice behavior is of great significance to urban traffic management and planning.The data obtained by traditional questionnaire cannot reflect the continuous travel process of travelers,but mining the travel characteristics of commuters from multi-source traffic information and analyzing the internal travel mechanism are important means to formulate effective traffic dredging strategies.Firstly,this paper preprocesses multi-source traffic data,and screens commuters according to their travel characteristics.By using the public transport operation data,the boarding and alighting stations and transfer behaviors are identified.The alighting station estimation model is established considering travel distance,station attraction,transfer capacity and the nature of surrounding land use,the transfer threshold is refined from the space-time dual dimension,and then the public transport travel chain is extracted.In addition,the effectiveness of this extraction method is verified according to the principle of basic balance between travel production and attraction of the station.The vehicle track is extracted from bayonet data,and the crossing time thresholds of every road segment by time period,so as to improve the classification accuracy of the single trip and obtain the private traffic travel chain.Then,according to the travel chain information,commuter’s travel mode is judged,key influencing factors are mined and characteristic values are extracted,and travel choice is combined with them as the model input.The random parameter logit model considering individual travel preferences is used to construct commuters’ travel mode choice model,and the marginal effects of each characteristic variable are calculated to analyze the influence and degree on travel mode choice,which provided a theoretical basis for the formulation of traffic guidance strategy.It has been proved that the fitting effect of this model is better than that of the multinomial logit model,and it can better explain commuters’ travel choice behavior.