Research Of Transportation Assignment Model Based On Ant Colony Algorithm

Transportation equilibrium assignment model is a mature and theoretical model. It is the base for transportation assignment problem. Because equilibrium assignment model has large number of variables and restriction conditions, it is difficult to solve it, which affects its application. Then researchers have put forward to many method to calculate the traffic flow such as Frank - Wolfe algorithm, Nerve Network, Genetic Algorithm and Fuzzy Algorithm, etc. But these methods can not actually and dynamically simulate the process while vehicles select the routes, so its result has some difference with the real world traffic conditions. To this point, my paper quotes Ant Colony Algorithm. This Algorithm not only has predominant stochastic search abilities, but also has self-adjustability and distributed trait, that's to say it has characteristics of colony cooperation, positive feedback and parallelism. And the processes of choosing route in Ant Colony Algorithm are similar with the process while vehicles select routes.Because data for Transportation equilibrium assignment model is very large, the paper use VB and the MapObjects which is 2-development component in GIS to establish database, realize Ant Colony Algorithm, Frank - Wolfe algorithm, and reflect results with special maps. Compare results of the two groups, together with non-equilibrium assignment results, we find that Ant Colony Algorithm is well applied and the results are reasonable, it provides a new method in solving transportation equilibrium assignment model.Through analyzing and studying, we get that the application with Ant Colony Algorithm is successful and valuable for practical use. It also provides a valid instrument to design rational traffic network.