基于增强蜂群算法的农产品多目标供应链优化  

Enhanced Bee Algorithm Based Agriculture Production Multi-target Supply Chain Optimization Approach

作　　者：彭剑[1] 

机构地区：[1]韶关学院信息科学与工程学院,广东韶关512000

出　　处：《控制工程》2016年第7期1123-1128,共6页Control Engineering of China

基　　金：广东省自然科学基金(2014A030313700);广东省科技计划项目(2013B070206076);广东省哲学社会科学项目(GD13XGL29);广东省普通高校特色创新项目(2014KTSCX171;2014WTSCX094);韶关市科技计划项目(2014CX/K252)

摘　　要：农产品的生产与供应对成本与供货时间要求极高,提出一种增强的蜂群算法来优化农产品基地的多基地、多目标供应链优化问题。首先,将蜂群中适应度值最高的地点选为目标地点,将其作为邻域搜索的输入参数,并为目标地点分配较多的搜索蜂,为非目标地点分配较少的搜索蜂;然后,经过一定次数的迭代搜索后,放弃其中改进不明显的地点,以此避免陷入局部最优并提高收敛速度;最终,将增强的蜂群算法结合农产品供应链进行实验与分析,获得了较好的优化效果。对比实验结果表明,算法获得较多的总成本与供货时间的帕累托最优解,可提供较多的供应链网络配置方案,同时,算法的鲁棒性与计算效率也具有优势。Abstract： The production and supply of the agriculture production need low cost and lead time, an improved bee colony algorithm is proposed to optimize multi-base and multi-production supply chain network of agriculture base. First, the fittest sites in the bee colony are selected as target sites, and set as the input parameters of the neighborhood search, the target sites are assigned more search bees and the rest sites are assigned less bees. Then, after several iterations, the sites which are not improved obviously are abandoned to prevent trapping in local optimal and improve the converge speed. Lastly, the multi-target supply chain network combined with enhanced bee colony algorithm is analyzed, and satisfactory effect is got. Experimental results show that the proposed approach gets more Pareto optimal solutions of total cost and lead time, and provides more configuration approaches of supply chain network, and the proposed algorithm has efficient computation and robustness.

关 键 词：蜂群算法 农产品基地 多目标供应链 随机邻居选择 供应链参数配置 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]