论文部分内容阅读
在假设各起始点的交通发生量和吸引量确定的情况下,研究不确定需求及不确定建设成本下的连续交通网络设计问题.利用鲁棒优化方法建立了一个不确定双层网络模型,设计了基于Frank-Wolfe法的情景局部搜索法.双层不确定模型中下层模型是需求属于一个简单有界闭集的用户均衡问题,上层模型是建设成本不确定的系统总阻抗最小的优化问题.通过数值算例分析,得到鲁棒优化下的不确定网络设计模型稳定性优于确定性设计模型的结论.
Under the assumption of the traffic volume and attraction quantity at each starting point, the design of continuous traffic network under uncertain demand and uncertain construction cost is studied. An uncertain two-layer network model is established by using robust optimization method. The local search method based on the Frank-Wolfe method is used.The lower-level model in the double-layer uncertain model is a user equilibrium problem in which the demand belongs to a simple bounded closed set, and the upper-level model is the optimization problem in which the total impedance of the system with undefined construction cost is minimized. Through the numerical example analysis, the conclusion that the stability of the uncertain network design model with robust optimization is superior to the deterministic design model is obtained.