Research On Hierarchical Optimization Of Public Traffic Network Based On Particle Swarm Optimization Algorithm With Mutation

With the development of economy and society, in our country, the development of urbanpublic transportation is relatively slow, and traffic congestion is more and more serious,which brings brought traffic misery to residents, especially public traffic network exists theproblems that layout of the line is unreasonable and the distribution of lines is unbalanced andscattered, which not only reduces the operating efficiency of public traffic network, but alsoseriously influences the service level of public traffic network. At present, most cities use nohierarchical method to achieve the optimization of conventional bus line networks, becausethe laying lines have unclear divisions of labor, a poor bridging and a lack of integrity, evenmore traffic blind areas are existed, so the method unable to improve the current unreasonablesituation of public traffic network. Based on the problems what the optimization of publictraffic network faces, using the hierarchical method optimize public traffic network has thepractical and important significance.This thesis mainly realizes the prediction of origin-destination (OD) trip distribution withthe combination of the improved particle swarm optimization (PSO) algorithm and greyprediction method. Hierarchical model is established based on the analysis of the advantage ofhierarchical optimization of public traffic network, a reasonable public traffic network isfound by using the improved PSO algorithm to solve each layer.Firstly, the continuous and the discrete PSO algorithm are improved. In the light of theshortcoming that the algorithm easily fall into local optimum in the process of solving, basedon the thought of improved strategy of dynamic index and genetic variation, the algorithm isimproved into PSO algorithm from combining the change of inertia weight with add mutationoperator, and the performance of the PSO algorithm with mutation and traditional improvedPSO algorithm are analyzed by using specific test function. Better convergence performanceof the PSO algorithm with mutation is tested and proved.Secondly, the advantage of using hierarchical method to optimize public traffic networkis analyzed, and the public traffic network of Lanzhou is choose as the research object, a grayparticle swarm algorithm with mutation combined prediction mode is established bycombining the PSO algorithm with mutation with grey prediction method, using this model,the prediction of OD trip distribution is implemented in MATLAB software, by comparingwith the predicted results of traditional growth counting method and gravity model method,the higher accuracy and the practicability of the combined prediction mode that is establishedin this thesis are proved by testing.Finally, this thesis uses random user equilibrium method to realize allocation of ODpassenger flow, based on it, the optimization models of main skeleton lines and inferior skeleton lines are built, each layer’s optimization constraints are determined, and the PSOalgorithm with mutation is used to find the solutions for each layer, the optimization andsetting of the main skeleton lines and the inferior skeleton lines are completed, the practical ofthe PSO algorithm with mutation and the effectiveness of hierarchical optimization areverified by analyzing and calculating the important indexes of the optimized public trafficnetwork. The hierarchical method can not only establish reasonable public traffic network, butalso improve the operating efficiency and service level of public traffic network.