摘要
中国部分山区高附加值生鲜农产品“最初一公里”已采用无人机运输模式,其中无人机集货中心的选址是提高时效性的关键,但各产区生鲜农产品产量的不确定性增加了选址的困难。针对该问题,引入多面体不确定集合刻画生鲜农产品产量的不确定性,考虑生鲜农产品分类及无人机集货中心分型,构建一个以总加权新鲜度最大化为目标的鲁棒选址优化模型。通过鲁棒对等理论将模型转化为混合整数规划,采用通用代数建模系统(general algebraic modeling systems,GAMS)中CPLEX求解器求解。案例分析表明:通过对四川雅江县野生菌无人机集货中心进行选址优化,运输量提高36%,新鲜度提高63%,单次运输时间节约84%,验证了模型的可行性和有效性,且在40组不确定水平和扰动比例随机组合的实验中,仅出现3种不同选址方案,说明模型具有较强的鲁棒性。
Drones have been applied to complete the“first mile”transportation of high value-added fresh agricultural products in some mountainous areas in China.The location of drone collection centers plays a pivotal role in improving timeliness.Yet difficulties increase because of the uncertainty of fresh produce output in different regions.To tackle this problem,a robust facility location optimization model was built driven by a need to maximize the aggregate of weighted freshness of produce,in the light of the sorts of agriculture products and the types of collection centers.The polyhedral uncertainty set was introduced to describe the yield uncertainty of fresh produce.The model was transformed into a mixed integer programming by robust equivalence theory,was solved by the CPLEX solver in general algebraic modeling systems(GAMS).The case analysis that the wild mushroom collection centers'optimization in Yajiang County,Sichuan,both shipments and freshness of produce rise,by 36%and 63%respectively.Yet whereas the single transit time decreases by 84%.The feasibility and effectiveness of the model have been demonstrated by the case study.Moreover,strong robustness of the model has been illustrated by 40 sets of experiments,leading to merely three facility solutions,randomly combined uncertainty levels with disturbance proportions.
作者
陈刚
贾晓朋
CHEN Gang;JIA Xiao-peng(School of Management,Guizhou University,Guiyang 550025,China;Karst Area Development Strategy Research Center,Guizhou University,Guiyang 550025,China)
出处
《科学技术与工程》
北大核心
2023年第28期12316-12323,共8页
Science Technology and Engineering
基金
国家自然科学基金(72261006,71761006)
贵州省教育科学规划项目(2022B017)。
关键词
设施选址
混合整数规划
生鲜农产品
无人机
鲁棒优化
facilities location
mixed-integer programming
fresh agricultural produce
drone
robust optimization
