The Research Of Traffic Assignment Problem On Urban Multi-modal Network

With the rapid development of national economy,the motorization level has been increasingly improved,which causes the increasingly serious urban traffic congestions.The critical reason for traffic congestion is the imbalance between urban traffic supply and demand.One of the effective way to alleviate traffic congestion is to reduce the imbalance between supply and demand via optimizing the supply.The core of traffic supply optimization lies in the realization of balance matching between urban multi-modal traffic demand and multi-modal traffic supply.Multi-modal supply refers to the comprehensive transportation system,including private car,public transportation,walking and bicycle.The goal of urban multi-modal traffic supply is to form a well-graded network with various modes,so as to guide the multi-modal demand to develop in the same direction,and finally achieve the balance between supply and demand.However,the multi-modal traffic network of most cities in China is not ideal.On the one hand,existing policies pay too much attention to the game between private cars and public transportation,and ignore the importance of other modes.The methods,such as bus priority and congestion pricing,all focus on only one or two modes in order to optimize the whole urban transportation system,which is certainly infeasible.The sudden rise and dramatic development of free-floating bike sharing just confirm this issue.On the other hand,the current field of transportation is in the stage of intelligent and connected vehicles.Emerging technologies,such as automated vehicle(AV)and connected and automated vehicle(CAV)have been attracting wide attentions from all of the society.With the unique performance of these intelligent vehicles,the changes of road capacity and traffic mode choice may have a subversive impact on the existing urban multi-modal transportation.Therefore,the new urban multi-modal traffic network considering intelligent and connected modes will play an important role in solving the imbalance of supply and demand and improving traffic congestions in the future.To this end,this paper conducted researches about the “traffic assignment problem on urban multi-modal traffic network” based on the key project of national natural science foundation “the modern urban multi-modal public transport system key technology,basic theory and performance”(51338003)and “the generalized transport hub city of multi-modal transport network collaborative planning theory and method”(51638004).Depended on the analysis of the basic method of traffic assignment on urban multi-modal traffic network,the paper conducted an in-depth study on the new multi-modal traffic network by considering the four aspects: free-floating bike sharing,AV,CAV,and dedicated lanes for CAV.First,the theoretical basis of traffic assignment on urban multi-modal traffic network was sorted out,providing a basis for the subsequent chapters.We mainly considered two modes,cars and public transportation,and introduced the user equilibrium(UE)model of private cars and the optimal strategy model of transit.Then,a Logit-based multi-modal network assignment model was developed.Last,the feasibility of the proposed basic model was verified based on the numerical examples of the Nine-Node and the Sioux-Falls networks.Second,we integrated both motorized and non-motorized traffic modes on one network,studied the traffic assignment model with the combined modal on a multi-modal network,and analyzed the impact of non-motorized mode on the results of multi-modal traffic assignment.Then,we considered that different users may have distinct mode choices,and developed a multi-class multi-modal traffic assignment model.Furthermore,based on the above models,we proposed a novel multi-modal assignment model integrating the free-floating bike sharing.Last,the feasibility of the proposed models were verified based on the numerical examples of the Nine-Node and the Sioux-Falls networks,and the influence of related factors such as the price of bike sharing was also analyzed.Third,we analyzed the impact of AV on multi-modal traffic network.The automatic detection and control of AV could reduce the burden of manual driving,improve safety and further enhance the traffic efficiency,which has a subversive impact on traditional traffic modes.We studied the road capacity of the mixed traffic flow consisting of AV and traditional regular vehicle(RV).Then,based on the new capacity function,we developed a new multi-modal traffic assignment model integrating the AV mode and proposed a corresponding algorithm to solve this problem.Furthermore,we developed a road pricing model based on the multi-modal network as well as a solving algorithm.Last,the feasibility of the proposed models and algorithms were verified based on the numerical examples of the Nine-Node and the SiouxFalls networks.Fourth,we analyzed the multi-modal traffic network considering CAV,AV as well as RV.We investigated the change of road capacity due to the environment of CAV and proposed a novel strategy to improve capacity,i.e.deploying roadside units to realize vehicle-toinfrastructure communication.Then,we developed a roadside units deployment mode to optimize the deployment of roadside units' position and number.Furthermore,we proposed a corresponding algorithm to solve the above problem.Last,the feasibility of the proposed model and algorithm were verified based on the numerical examples of the Toy,the Nine-Node and the South-Florida networks.Finally,we analyzed the issue of dedicated CAV lane based on multi-modal network consisting of CAV and RV.We investigated the road capacity of dedicated CAV lane,developed the CAV lane deployment model on multi-modal traffic network,and proposed the corresponding algorithm for solving this problem.Then,we developed a novel mixed deployment model,which considers both the dedicated CAV lane and roadside units.A solving algorithm was also proposed for this problem.Last,the feasibility of the proposed models and algorithms were verified based on the numerical examples of the Nine-Node networks.