Abstract: The urban road network design problem deals with how to add or improve some links on an existing traffic network using quantitative analysis method. A bi-level programming model of urban mixed network design based on genetic algorithm is presented in this paper, which can calculate the optimal transportation investment decision of link improvements or link additions. Using the model and algorithm, a simulation test has been made in an urban road network design. The results indicate the model and algorithm are feasible, and they have provided a basis for urban road network design problem.

Key words: mixed network design, bi-level programming, genetic algorithm

摘要： 城市路网设计问题就是研究如何用定量的方法在已有交通网络上添加或扩容某些路段的问题。本文提出一种基于遗传算法的城市混合型路网设计的双层优化模型，可求出最优的用于道路网新建或改善的交通建设投资决策方案，并利用一个算例进行仿真试验，结果表明，该模型和算法是可行的，可为城市路网设计提供借鉴。

关键词: 混合型路网设计, 双层规划模型, 遗传算法