Research On Traffic Optimization Problem In Special Situations Of City

With the rise of China’s economy and international status,more and more international competitions and large-scale activities are held in domestic cities.As a result,the traffic flow in this region has increased dramatically,which makes it difficult to evacuate people in a short time.So,study the method of emergency evacuation and the city traffic control,in order to audience who remain after a large-scale activity can be evacuated in a short time.Reducing traffic impact caused by the large-scale activity.Improving the efficiency of urban traffic evacuation under special situations has important practical significance for the traffic management of city.This thesis has mainly researched on traffic evacuation of large-scale activity.The main contents of this thesis include:1.This thesis introduces the definition and classification of urban special situations base on studying relevant data.Analyzing traffic features of large-scale activities,given a method for computing the scope of its impact.This thesis presents the calculation method of evacuation demand.2.According to the features of city road network under special situations,this thesis improves the Astar algorithm and proposes a new algorithm-AAstar algorithm.Through the experimental comparison and analyzing of Dijkstra,Astar and AAstar algorithm,it shows that AAstar has obvious advantages in the route planning of dynamic road network.3.On the basis of summarizing the vehicle scheduling problem and its solving algorithm,aiming at the features of large-scale activities,we establish the mathematical model of vehicle evacuation scheduling problem.Particle swarm optimization algorithm is characterized by outstanding global optimization ability,low dependence on the problem and slow convergence speed in the late iteration.Combined with the strong local search ability of bacterial foraging algorithm,this thesis proposes a particle swarm optimization algorithm based on bacterial foraging.We choose three Benchmark functions to verify both convergence speed and convergence accuracy,compare with several improved particle swarm optimization algorithms,the results show the algorithm has a distinct advantage in solving complex nonlinear functions.At the same time,we apply the new algorithm to solve the vehicle evacuation scheduling problem of large-scale activities,it has proved the superiority of the BF-PSO algorithm.Finally,we combine the background of large-scale activities,establish an evacuation emergency model integrating route selection and vehicle evacuation scheduling.The improved algorithm is used to carry out the model.The experimental results show the algorithm proposed in this thesis has some practicability and provides a decision-making basis for urban emergency evacuation traffic planning and management.