Abstract: The paper presents the optimal model and algorithm of mixed transportation network design problem considering the sustainable development. The decision-making variables of road level status are used to make the mixed network discrete, and the bi-level programming model is used to describe the problem. The objective of the upper model is to minimize the total investment and emission of cars. The constraint conditions are set as the expropriation of the transportation network and the degree of traffic congestion of the roads. The determinate user equilibrium assignment is selected as the lower level model. The simulated annealing algorithm and the path-based gradient project algorithm are used to solve the upper and lower model respectively based on the attribute of the model. The results of numerical example show that the congestion is relaxed observably and the emission of cars are decreased substantially in the optimized network, which indicates the model and algorithm are available for road network design considering sustainable development.

Key words: traffic engineering, sustainable development, transportation network design, bi-level programming model

摘要： 研究了考虑可持续发展的混合型交通网络设计问题的优化模型和算法. 利用双层规划模型描述问题，其中上层模型是以方案总投资费用和汽车废气排放量最小化为优化目标，并受占用土地规模和路段负荷度的约束，下层模型为确定型的交通网络平衡配流模型. 使用路段等级决策变量将问题离散化，并基于模拟退火算法和基于路径的梯度投影算法进行模型求解. 算例计算表明，优化后的交通网络不仅拥堵状况得到显著缓解，而且路网内汽车废气排放量下降幅度也非常明显. 这都证明本文所提出的双层规划模型和求解算法是一种研究可持续发展环境下交通网络设计问题的有效方法.

关键词: 交通工程, 可持续发展, 交通网络设计, 双层规划模型