Equilibrium Flow Based System Optimization Model And Algorithm For Urban Transit Network

With the process of urbanization accelerating, the level of urbanization rising and the urban population increasing, urban travel demand substantially increases. At the same time, the rapid development of the car on the one hand brings the convenient travel, on the other hand, a sharp rise in the number also greatly increases the pressure on the existing road network. The planning and construction of the existing road network is failed to meet the pace of economic development, which brings about urban traffic issue. Urban traffic congestion has become a stumbling block to the cities" development, which must be addressed now. Urban transit system has features of huge capacity, convenience and so on. Vigorously developing urban transit has become the first choice to solve the urban traffic issue.Around the problems of the optimal design of the urban transit network, the research work of this paper is as follows:Firstly, urban transit passenger demand is an important basis for the construction of urban transit system. The reasonable forecasting and simulation of the passengers’ distribution in the transit network helps lay the passenger data base of the transit system optimization design. Because urban transit passenger flow is diverse, varied, complex, building passengers’assignment model reasonably affects the effect of the simulation of the passenger distribution in the network directly. Based on this, this article presented a stochastic user equilibrium model for the urban transit assignment problem considering the affection of the transfer factor, with the transfer costs and transfer times considered. The stochastic user equilibrium model simulated passengers" travel route choice behavior and passengers’distribution in the network better.Secondly, the departure frequency optimization is one of the most important work in the urban transit network optimization design, which not only affects the efficiency of urban transit services, but also directly related to the effectiveness of the bus operating companies. Based on the interests of both passengers and bus companies, this paper established the bi-level optimization model of urban transit system departure frequency. The minimum of passengers’travel total cost, the maximize of the bus operating companies income is the upper optimization objectives, and the stochastic equilibrium assignment model considering the transfer costs is the lower optimization goal. The improved genetic algorithm was used to solve this problem. The improved public transit departure frequency is more in line with the interests of passengers and bus companies.Thirdly, urban transit network plan subject to the multifaceted conditions and factors. This paper introduced the principles, objectives and influencing factors of the urban transit network optimization design. On this basis, considering the line length limitation in the bus system, the minimum passenger flow restrictions, non-linear limitation, the cross section flow limitation, site distance limitation, the departure frequency limitation and so on, this paper established urban transit network optimization design model with the maximum of the passengers directly arriving rate and the bus operating companies income being the goal for the upper model, the stochastic user equilibrium assignment as the lower model. It made the interests of passengers and operators taken into account as much as possible.Finally, Genetic algorithms has obvious advantages in solving optimization problems. This paper made some improvement on Genetic algorithms. For urban transit network optimization problem, this paper made a combination of the genetic algorithms and the other optimization algorithm (simulated annealing algorithms), avoiding the shortcomings of the genetic algorithms in the solution process effectively. The improved hybrid genetic algorithms has better convergence and solution efficiency.