修建成本不确定的交通网络设计问题

Network Design Problem with Uncertain Construction Costs

交通网络设计问题是交通规划理论的一个重要组成部分,即在资金有限且考虑出行者决策行为的情况下,制定最优投资策略.由于人工费、材料费和使用费等的不确定性,路段的修建成本存在不确定性.本文通过改进预算投资约束,应用鲁棒优化的方法同时考虑出行者的路径选择行为,建立路段修建成本不确定的交通网络设计的鲁棒模型,并利用基于割约束的混合整数线性规划算法求解此模型,进而得到一个受修建成本扰动较小的鲁棒最优解.通过算例表明,在修建成本不确定的交通网络设计中,本文提出的鲁棒优化方法可以得到比传统确定性问题更加可靠的解.

The transportation network design problem is an important component of transportation planning theory. With consideration of limited financial resources and travelers＇ decision-making behavior, it aims to develop the optimal investment strategy. In this paper, a novel network design problem under uncertain construction costs is proposed. Based on the robust optimization method ,the budget investment constraints are improved to overcome the limitations of the existing approaches. A global optimization algorithm based on a limited generated method is developed for solving the mixed integer linear programming. A series of comprehensive computational experiments show that in transportation network design problem, the proposed robust optimization method can provide more reliable solutions than the traditional deterministic optimization methods.