Research On Transportation Demand Response Mechanism Under Mobility-as-a-Service Model

With the accelerated urbanization,a series of problems such as traffic congestion,environmental pollution and resource waste have emerged in cities,and the imbalance between the supply and demand of various modes of transportation is one of the important causes.The supply of transportation resources during peak periods is much lower than the demand for travel,thus aggravating traffic congestion and delays;while the supply of transportation resources during flat periods is higher than the demand for travel,resulting in a waste of resources.How to balance the supply and demand of various modes of transportation capacity and provide higher quality transportation services has become an urgent research problem to be solved.The transportation demand response model can formulate plans and allocate resources from the perspective of the demand side,and then reasonably solve the problem of imbalance between the supply and demand of traffic capacity.The traditional demand response model has a single means of service,and the service targets are limited by vehicle capacity,cost and other factors,which cannot meet the needs of all travelers.Mobility as a Service(Maa S)is a new type of transportation service system proposed in recent years,which can integrate all kinds of transportation resources from the perspective of transportation demand and provide customized,more flexible and higher quality one-stop travel services for travelers.Therefore,it is of practical significance to study the transportation demand response mechanism under the Maa S model.This study constructs a transportation demand response model under Maa S model in order to respond to the traffic and travel demand under Maas system,improve traffic efficiency and refine its system framework design.A spatiotemporal preference clustering algorithm is designed to improve the response rate and achieve full demand coverage by considering travel preferences.The specific work of this paper includes:(1)Describes the current state of development and key attributes of transportation demand response and Maa S systems.Considering the technical aspects and service attributes of Maa S system,a transportation demand response model under Maa S model is constructed.Collection of reservation demand point data through a network platform built by transportation companies to provide travel services combining fixed,semi-fixed and flexible transportation modes.It also coordinates and allocates transportation resources and improves the quality of travel services based on the spatial and temporal distribution of travel demand and the priority order of travel preferences.(2)Considering the spatiotemporal distribution of travel demand and travel preference factors,a spatiotemporal preference clustering algorithm is designed to meet the clustered response of reservation demand points.Improved time preference hierarchical clustering based on Maa S system and travel preference as a perceptual factor.The spatial DBSCAN clustering is improved by combining the travel preference priority order,and the contour coefficients are used as parameter update indicators,and then the reservation demand points with convergent spatio-temporal preferences are obtained and responded to.Combined with travel preference and preference priority order,the DBSCAN clustering of spatial preference is improved,and the contour coefficient is used as the parameter update index to obtain the reservation demand points with the convergence of spatiotemporal preference and respond.At the same time,the demand response model is used to coordinate the supply of various types of transportation resources and allocate flexible routes to serve spatial and temporal discrete points to achieve full coverage of demand.(3)A simulation database is constructed to verify the feasibility of the models and algorithms,to quantitatively analyze their improvement efficiency and to discuss applicability.Simulation and research results show that the model and algorithm can effectively cluster response reservation demand points.The average value of its demand response rate is 93%when the number of reservation demand points is 60.Compared to the hierarchical clustering and DBSCAN clustering algorithms,the improvement is 15% and 6.7%,respectively.The improvement is 8.3% compared to the customized bus model.The model and algorithm are applicable to areas where public transportation is dominant in static demand response mode,and can effectively improve the efficiency of traffic demand response and service quality.