Urban Dynamic Traffic Assignment Model And Algorithm Based On Improved Cell Transmission Model

With the rapid development of urban economy and scale,traffic congestion,environmental pollution and other issues have become extremely serious urban disease.Developed countries have invested heavily in the research of Intelligent Traffic System(ITS)to alleviate these urban disease.ITS can exploit the potential of existing transportation facilities,improve transportation efficiency,alleviate congestion and reduce pollution,thus achieving great social and economic benefits.Dynamic traffic assignment(DTA)assign time-varying traffic trips reasonably to the network according to certain rules.Different from the static traffic assignment aiming at traffic network planning,dynamic traffic assignment is a demand forecasting method targeted at road traffic flow and traffic management and control of timeliness.It is one of the core technologies of ITS.Constructing a dynamic traffic assignment model that can clearly reflect the traffic flow changes in the space-time dimensions and researching a solution algorithm that can deal with the huge amount of computation brought by large-scale network has become a hot and difficult point in this field.Therefore,this paper will start with the impedance function and algorithm of dynamic traffic assignment model,improve the accuracy and applicability of dynamic traffic assignment model and strengthen the calculation efficiency of the algorithm,finally lay a theoretical foundation for the practical application of the model.Firstly,the development and classification of the dynamic traffic assignment model are elaborated in detail,including the development and application of its impedance function and algorithm.Then,the insufficiency of current research is analyzed and the research goal of this paper is clarified.Subsequently,on the basis of the theory of Cell Transmission Model(CTM),the improved Variable CTM(VCTM)is selected to improve the VCTM of urban road,combining with the traffic flow hysteresis characteristics in urban roads.Secondly,considering the operation law of traffic flow in urban nodes,the node cell partition method is proposed,and the rules of confluence and diversion of node joint section are improved,which are combined into node VCTM.Integrating road and node models into the urban network VCTM(UN-VCTM).Finally,based on UN-VCTM,the existing fully discretized method for calculating the actual impedance of roads is improved by the composite method of discrete and continuous calculation.Furthermore,on the basis of Ant Colony System(ACS),considering the actual route selection characteristics of travelers under the background of popularization of intelligent facilities,UN-VCTM is substituted as impedance function into Dynamic User Optimization(DUO)model to replace the original BPR function.Then the state transition probability and pheromone updating rules of ACS are improved,and an adaptive feedback mechanism based on road condition information is designed,and then an adaptive ACS is proposed.Finally,two different case networks are selected for simulation and verification.Firstly,the MATLAB program of UN-VCTM and the composite impedance calculation method are designed.At the same time,case one is also simulated and analyzed by VISSIM.The output results show that the model and method described in this paper have the advantages of simple simulation and high accuracy compared with VISSIM.Then,on the basis of case two,using adaptive ACS and two comparison algorithms to solve the dynamic user optimal model,the analysis of network assignment results shows that the adaptive ACS assignment results described in this paper are more balanced and can dynamically adjust the impedance of each road so as to achieve the lowest overall impedance,which is more realistic.