中文摘要：

考虑综合交通枢纽场站协调优化布局，在传统客运枢纽规划模型中增加用户对运输模式的选择与枢纽中转能力两个约束条件，提出基于容量限制与运输模式选择的综合客运多枢纽布局非线性整数规划模型，设计了改进的遗传算法对其进行求解。应用LINGO软件对布局优化模型进行有效性检验，对8节点Solomon标准测试数据进行计算，得出LINGO平均运算时间为5102S，最优成本为1899782元；遗传算法MATLAB编程平均运算时间为59S，最优成本为1948796元。对50节点数据进行运算，平均运算时间为569s，最优成本为8497602元；使用25节点规模的AP数据集合，取枢纽数量为3时得出的最优成本为154932元，应用传统枢纽规划模型进行求解，平均运算时间为607S，最优成本为155098元，比经典算法降低了166元。可见，与传统枢纽规划模型相比，该模型与算法最优成本更少，说明改进的枢纽布局优化模型有效。

英文摘要：

Considering coordinated layout of comprehensive transportation terminal, this paper increases two con-straint conditions for transportation pattern and hub transfer ability in traditional model of passenger terminal planning, puts forward a nonlinear integer programming of comprehensive passenger transport multiple hubs based on capacity limi- tation and transportation pattern, and designs an improved genetic algorithm. LINGO software is applied to the model ef- fectiveness test of optimized layout to cumulate the standard data of 8 nodes Solomon, the average operation time of LIN-GO is about 5 102 s and the optimum cost is 1 899 782 Yuan; the average operation time of genetic algorithm program- ming is about 59 s and the optimum cost is 1 948 796 Yuan. The average operation time is about 569 s and the optimum cost is 8 497 602 Yuan if cumulating 50 nodes data; the optimum cost is 154 932 Yuan if using AP data set of 25 nodes and taking 3 hubs, the average operation time is about 607 s and the optimum cost is 155 098 Yuan through traditional hub planning model, while the optimum cost reduces 166 Yuan when compared to the classical algorithm. So the optimi-zation model of developed joint terminal layout has faster solving speed and less optimal cost than traditional model of joint terminal planning.