Synthetically Improved Genetic Algorithm in Public Traffic Dispatch System

On the basis of the bus scheduling problems, using stochastic optimization theory and nonlinear programming theory and so on, the bus scheduling and the situation of bus arrangement of Shijiazhuang Hebei province were analyzed. Study on the difference of this process and the improvement between the actual situation and the optimum scheme in theory. Because there are many of random factors among the bus scheduling in the practical environment, on the base of nonlinear programming theory, the random analysis theory was used. A stochastic nonlinear programming model and solutions to them were proposed, objectively the defects of poor objectivity of nonlinear programming theory was made up. At the same time the paper introduces a synthetically improved genetic algorithm which utilizes the advantage of other meliorated genetic algorithm and synthesizes them. Therefore the synthetically improved genetic algorithm's capability is enhanced and the result of model which is solved using it is excellent in the solution space. So the abuse of traditional model was avoided. It considered the dual interests among passenger and the bus company. And it researches a balance point among them.