新型城镇化模式下的离散交通网络设计模型与算法

A Model of Discrete Transport Network Design and Its Algorithm under New Urbanization Mode

为使交通规划与生态、资源相协调,基于新型城镇化发展理念,提出了区域差别化排放约束和区域差别化道路土地资源约束,结合Pareto最优解思想,构建了离散交通网络设计双层规划模型。其中上层模型以系统阻抗、投资费用为优化目标,以区域差别化约束为约束条件;下层模型为固定需求下的用户平衡配流模型。在第二代非劣排序遗传算法(NSGAII)基础上设计了模型的求解算法,同时为验证算法是否出现早熟收敛,设计了第k小距离策略代替拥挤距离策略的验证算法,并在Matlab平台上开发了相应的算法程序。在经典Nguyen-Dupuis网络上,求取了模型的Pareto最优解,并验证了设计算法的有效性。为分析约束强度区域范围变化对决策值的影响,进行了区域划分的鲁棒性分析,并得到了资金较充裕与不足情况下的目标决策值变化趋势。

In order to coordinate transport planning with ecology and resource,based on the development idea of new urbanization,the regional differentiated constraint of discharge and the road land resource constraint are presented,combing with the idea of the optimal Pareto,a bi-level programming model of discrete transport network design is established. In the upper-level programming model,the system impedance and investment cost are used as the optimization objective,the regional differentiated constraint is used as the constraint condition; in the lower-level programming model,the user equilibrium assignment model under the fixed traffic demand is used. Based on the NSGAII,the solving algorithm is designed for the model,moreover,in order to test the premature convergence performance of the algorithm,the tested algorithm that replacing the crowding distance strategy with the kth small distance strategy is designed, and the corresponding program is developed using Matlab. The optimal Pareto solution of the proposed model is got on the classical Nguyen-Dupuis network,and the designed algorithm is tested. To analyze the influence of the regional scope of the constraint strength on the decision value,the robustness analysis of region partition is conducted,and the change trends of objective decision values under the condition of sufficient funds and insufficient funds are got.