考虑网络韧性的三级轴辐式物流网络模型及算法研究

为了实现城市物流运作的集约化，并尽可能降低因突发事件导致的物流供需中断，以物流节点定位和通道布局为落脚点，研究了三级轴辐式物流网络的多目标规划。考虑轴辐式网络的结构约束、多级节点间的连通关系限制，以多级轴辐点选址、轴辐点间的连通关系及强度为决策变量，以网络建设成本最小和网络韧性度最大为目标函数，构建三级轴辐式物流网络模型。结合模型的特点设计了具有双层结构的遗传算法，其中外层为三段编码结构的小生境Pareto遗传算法，用于求解多目标规划，内层为0-1编码的遗传算法，用于求解网络韧性度。以某区域14个节点城市作为案例，验证模型及算法的有效性，并将优化结果与两类传统的单目标物流网络进行了比较。研究结果表明：双层遗传算法能够有效求解三级轴辐式物流网络规划模型；选取案例计算得到的Pareto最优解参与比较分析，优化结果与传统三级轴辐式物流网络相比，网络韧性度提高了6倍，但网络建设成本也随之增加；与单纯以韧性度最大化为优化目标的物流网络相比，网络建设成本降低了7倍，网络韧性度也随之降低。可见提出的双目标三级轴辐式物流网络模型实现了网络建设成本和网络韧性度间的均衡，对于追求具备一定抗毁性能的轴辐式物流网络规划人员，具有参考和应用价值。

Research on Three Level Hub and Spoke Logistics Network Model and Algorithm Considering Network Tenacity

In order to realize the intensification of urban logistics operation, at the same time, the interruption of logistics supply and demand caused by emergencies shall be reduced as much as possible.Take logistics node positioning and channel layout as the foothold, the multi-objective programming of three-level hub and spoke logistics network is studied. Considering the structural constraints of hub and spoke network and the connectivity constraints between multi-level nodes, taking the location of multi-level hub and spoke points, the connectivity and strength between hub and spoke points as decision variables, and taking the minimum network construction cost and the maximum network tenacity as objective functions, a three-level hub and spoke logistics network model is constructed. Combined with the characteristics of the model, a genetic algorithm with double-layer structure is designed, in which the outer layer is a niche Pareto genetic algorithm with three-stage coding structure for solving multi-objective programming, and the inner layer is a genetic algorithm with 0-1 coding for solving network tenacity. Taking 14 node cities in a region as an example, the effectiveness of the model and algorithm is verified, and the optimization results are compared with two kinds of traditional single objective logistics networks. The results show that bilevel genetic algorithm can effectively solve the three-level hub and spoke logistics network planning model. Select the Pareto optimal solution obtained by case calculation to participate in the comparative analysis. Compared with the traditional three-stage hub and spoke logistics network, the optimization results have improved the network toughness by 6 times, but will bring a certain increase in network construction costs. Compared with the logistics network whose optimization goal is to maximize the flexibility, the network construction cost is reduced by 7 times, and the network flexibility is also reduced. It can be seen that the proposed dual objective three-stage hub and spoke logistics network model achieves the balance between network construction cost and network toughness, which has reference and application value for planners pursuing hub and spoke logistics network with certain invulnerability.