基于多种群进化小生境遗传算法的电力立体式仓储仓位多目标优化配置方法  

Multi Objective Optimization Configuration Method for Power Three-dimensional Storage Space Based on Multi-population Evolutionary Niche Genetic Algorithm

作　　者：苏明 王增超 郝锋 李科锋 吴俊兴 SU Ming;WANG Zengchao;HAO Feng;LI Kefeng;WU Junxing(State Grid Hebei Electric Power Co.,Ltd.Materials Branch,Shijiazhuang 050021,China;.2.NARI Nanjing Control System Co.,Ltd.,Nanjing 210061,China)

机构地区：[1]国网河北省电力有限公司物资分公司,河北石家庄050021 [2]国电南瑞南京控制系统有限公司,江苏南京210061

出　　处：《河北电力技术》2024年第5期48-54,共7页Hebei Electric Power

基　　金：国网河北省电力有限公司科技项目(kj2021-053)。

摘　　要：针对传统电力立体式仓储仓位配置中存在关联性和灵活性不足的问题,提出一种基于多种群进化小生境遗传算法的电力立体式仓储仓位优化配置方法。首先,对物资进行关联性分析,通过聚类方法将物资科学分类,以提高仓储操作的效率和准确性。然后,基于物资的聚类结果和特性设定约束条件,以物资出入库时间和类内物资距离最小化为目标,构建了仓位优化配置目标函数。最后,引入多种群进化小生境遗传算法进行求解,在仓储仓位搜索空间中更好地保持可选配置方案的多样性,避免过早收敛,在全局范围内寻找最优解,实现电力立体式仓储仓位的优化配置。实验结果表明,所提方法能够提高仓储效率、满足物资管理需求,且在具有更高优化程度的同时保持较高的关联性与灵活性。A multi-population evolutionary niche genetic algorithm-based optimization configuration method is proposed to address the challenges of correlation and insufficient flexibility in traditional electric power three-dimensional storage space allocation.First,conduct correlation analysis is performed on materials,and clustering methods are used for scientific classification,improving the efficiency and accuracy of warehouse operations.Based on the clustering results and material characteristics,constraints are set,and an objective function is constructed with the aim of minimizing material entry/exit time and intra-class material movement distance.The introduction of a multiple-population evolutionary niche genetic algorithm helps maintain diversity in storage bin configuration solutions,avoids premature convergence,and enables global optimal solution search.The experimental results demonstrate that this method significantly enhances warehouse efficiency,meets material management needs,and achieves a high degree of optimization with correlation and flexibility.

关 键 词：聚类分析 优化配置模型 多种群进化小生境 遗传算法 模型求解 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置]