Urban Dynamic Traffic Assignment Model Based On Improved Genetic-Ant Colony Algorithm

Urban traffic congestion has become a common problem in every city.Traffic congestion causes slow driving speed and increased travel time,as well as serious environmental pollution and economic losses.Intelligent transportation system can reasonably allocate travel demand according to the existing traffic facilities,improve travel efficiency and alleviate traffic congestion on the road network.So this thesis studies the dynamic traffic flow assignment model DTA of ITS core part and the corresponding solution algorithm,so as to reduce travel costs,enhance road traffic efficiency and provide solutions for alleviating urban traffic congestion.Therefore,the research work of this paper is of great practical significance.Firstly,the thesis analyzes the research progress of dynamic traffic flow assignment model and solving algorithm at home and abroad,and determines the research objectives and contents of the thesis.Then,The theory and basic characteristics of dynamic traffic flow assignment are introduced,two key problems of dynamic traffic flow assignment model,the equation of state and the objective function,are analyzed and discussed.Based on the continuous time equation of road section state and the objective function of dynamic user optimization,a dynamic traffic flow assignment model is established to describe the real road network.Furthermore,the genetic algorithm and ant colony algorithm used to solve the dynamic traffic assignment model are studied and analyzed.The basic principles and characteristics of the two algorithms are introduced,and TSP30 experiments are carried out to solve the two basic algorithms.The simulation results are analyzed and compared,and the parameter setting is determined,which lays a foundation for the fusion algorithm.Next,the thesis analyzes the fusion mode and the key problems of the convergence of the two algorithms,and determines the series fusion mode and the fusion time.In order to maintain the diversity of the population and keep the high quality genes from being destroyed,the reinsertion of the parent generation was added into the genetic operation of the basic genetic algorithm,so as to optimize and improve it.Ant colony algorithm is optimized by combining genetic algorithm preprocessing with pheromone restriction and pheromone update rules of maximum and minimum ant system algorithm.Combining the improved genetic algorithm and ant colony algorithm,a new improved genetic-ant colony algorithm was established,and the TSP was also solved by the new algorithm,and the conclusion was drawn that the improved genetic-ant colony algorithm is efficient,has high solution quality,s Tab and better performanceFinally,Nguyen-Dupius road network was selected as an example,and Dynamic user optimization model was solved by improved genetic-ant colony algorithm and ant colony algorithm.Through the simulation results,the improved genetic-ant colony algorithm reduces the proportion of road sections that occupy more than 80% of the road space and avoids the traffic congestion caused by excessive traffic flow in some road sections.At the same time,the utilization rate of sections with a space occupancy of less than 80% has been improved to make full use of road network resources and improve traffic efficiency.And according to the objective function value of the DUO model solved by the two algorithms,the optimal solution of the improved genetic-ant colony algorithm is better,and the variation range of each generation solution is smaller and more stable.