Abstract: To analyze the headway optimization theory about urban public transport.The optimization of transit operation parameters such as headway is a very important task in APTS.The paper establishes a new model based on the maximum of the overall social benefit.Because so many variables are involved in the model，the traditional optimization method is infeasible to solve it.This paper proposes a new method to optimize those parameters using Genetic Algorithm（GA）.The method mentioned in this paper is applied to the actual urban transit operations of two public transportation routes in Changchun City.

Key words: genetic algorithm, intelligent public transport dispatching, headway

摘要： 在分析公共交通线网发车间隔优化理论的基础上，从公共交通线网间发车间隔对公交网络系统进行优化出发，提出了基于社会总体效益最优的思想，构建了公共交通线网发车间隔优化模型，从而解决了智能公交中的线网间智能调度的难题，一种新的生物进化算法遗传算法被引入到问题的求解过程中。模型和算法应用于长春市的典型线路，获得了较好的优化结果。

关键词: 遗传算法, 智能公交调度, 发车频率