基于复合种群SIR模型的应急防疫物资仓库选址

Warehouse site selection of emergency epidemic prevention materials based on metapopulation SIR model

为提高对疫情的应急响应与控制水平,研究应急防疫物资储备仓库选址方法。首先,综合考虑城市网络中人口及城市间的人流量因素,利用复合种群SIR模型预测应急防疫物资需求量,并以2017年华北地区城市流感数据验证预测准确性;然后,通过栅格化选址区域生成初始解空间,以时效性优先为原则,构建基于P中值模型的大规模区域内储备仓库选址模型,并以加权运输距离最小为目标,设计结合重心法的精英保留遗传算法求解模型;最后,以华北地区防疫物资仓库建设为试验案例,使用真实运输距离数据验证模型和算法的有效性。结果表明:在大规模区域内仓库候选位置未知的情况下,该模型和求解算法能够保证选址方案的合理性和计算敏捷性,在仓库数量有限的条件下满足疫情爆发时的应急防疫物资供应需求。

In order to improve the emergency response and control level of the epidemic situation,the location method of emergency epidemic prevention material reserve warehouse was studied.Firstly,considering the population in the urban network and the flow of people between cities,the composite population SIR model was used to predict the demand for emergency epidemic prevention materials,and the accuracy of the prediction was verified by the urban influenza data in North China in 2017.Then,the initial solution space was generated by rasterizing location area.Based on the principle of timeliness priority,a large-scale regional reserve warehouse location model based on p-median model was constructed.With the goal of minimizing the weighted transportation distance,an elite retention genetic algorithm combined with the center of gravity method was designed to solve the model.Finally,taking the construction of epidemic prevention material warehouse in North China as an experimental case,the validity of the model and algorithm was verified by using real transportation distance data.The results show that when the candidate location of the warehouse in a large-scale area is unknown,the model and the algorithm can ensure the rationality and computational agility of the location scheme,and meet the supply demand of emergency epidemic prevention materials under the condition of limited number of warehouses.