Influence Of Multilevel Factors On Spatio-temporal Evolution Of Online Car-hailing Demand From The Perspective Of New Restriction Policy

Identifying and understanding factors that influence the car-hailing demand is essential for online hailing systems to improve the quality of service.In this thesis,a two-level growth model(GM)based on the city circle layer division and grid-based division was proposed to identify potential multi-level factors that may affect online car-hailing service demand.By using the massive datasets from Di Di Chuxing,Inc.,including both Di Di Taxi and Di Di Express services data,the model was able to track the market changes before and after Shanghai’s new operation policy at the same time.The fluctuations in order number at different city circle zones were analysed to assess the competition and mutual complementarity between Di Di Taxi and Di Di Express.By bringing the relative market share of Di Di Express into the model,the difference in characteristics between the two types of services is further revealed.Based on the existing research,this study further confirmed the significant positive correlation between the demand for online car-hailing service and the built environment factors,such as commercial/residential land use,public transport service level and other positive effects on demand formation.It is also found that there is a stronger correlation between metro service and the car-hailing demand than the conventional bus.In terms of meteorological factors,the study found that Di Di Express occupied a higher market share in rainy days,and attributes it to the adaptive pricing mechanism and flexible working hours for the drivers.In addition,the study also revealed the difference in the effect of the new operation policy in different urban circle layers,and put forward the "squeezing effect" on the express business in the central city.The research results provide reference for the operation and management of the online car-hailing service.At the same time,it provides important demonstration for urban planning and management,especially the formulation of relevant policies and the prediction of policy effects in the future.