基于遗传算法的河南省应急物流中心选址研究
Genetic Algorithm-Based Site Selection Study of Emergency Logistics Center in Henan Province

以河南省为例，通过分阶段递进式的方法进行初步选址和最终选址。首先根据影响应急物流能力的主要因素形成指标数据，将其带入熵权评价模型，得到河南18个地区应急物流能力评价指数排名，采用Ward聚类法将18个地区分为4类，并剔除4个地区。再将剩余14个地区设为物资供应点，全省109个节点为需求点，通过遗传算法求解满足需求加权距离最小化且带有容量限制和时间窗约束的选址模型，通过多次实验将数据带入模型得到最终理想选址方案，既能提高应急物流中心的响应速度，也能减少建设成本，最大程度的保证应急物流中心选址的合理性。
In recent years, all kinds of natural disasters have been high, seriously affecting the normal order of society and the safety of people’s property. The state attaches great importance to the comprehensive management of natural disasters and emergency response capacity building, and vigorously carries out all emergency management work. For the emergency logistics center as an important part of the emergency logistics system and emergency logistics network and key nodes, its scientific and reasonable location planning can improve rescue efficiency, reduce emergency costs, reduce losses from natural disasters. Therefore, at the pre-disaster stage, how to build an emergency logistics center at a key location in a region and reserve materials in advance, so that emergency materials can be delivered in a timely and efficient manner to the point of demand for disaster relief after the disaster has occurred will become very meaningful and critical, which is precisely the focus of this paper. Taking Henan Province as an example, the preliminary site selection and final site selection were carried out through a phased progressive approach. Firstly, the index data is formed according to the main factors affecting the emergency logistics capacity, which is brought into the entropy power evaluation model to get the ranking of the emergency logistics capacity evaluation index of 18 regions in Henan, and the Ward clustering method is used to divide the 18 regions into 4 categories and eliminate 4 regions. Then, the remaining 14 regions are set as material supply points, and 109 nodes in the province are demand points. The location model with the minimum weighted distance of demand and with capacity constraints and time window constraints is solved by genetic algorithm. The final ideal location scheme is obtained by bringing data into the model through multiple experiments, which can not only improve the response speed of the emergency logistics center, but also reduce the construction cost, to ensure the rationality of emergency logistics center location to the greatest extent.