【作者】 王丽；

【导师】 包菊芳； 董锐；

【作者基本信息】 安徽工业大学 ， 物流工程， 2021， 硕士

【摘要】 随着电子商务的迅速发展,仓储中心自动化程度越来越高,针对一般大型仓储中心内快消品存储量较大、种类较多的情况,运用AGV(Automated Guided Vehicle)移动货架的“货到人”拣选作业模式应运而生并逐渐被众多大型电商仓储中心采纳使用。运用AGV搬运移动立体货架取代工作人员的移动,搬运一次货架甚至可以实现多个SKU的同时拣选,极大地提高了订单拣选效率、减少了人力资源成本。同时伴随着客户订单需求的多样化改变,传统单一的随机储位、分类储位策略等已经不能满足货位规划的需求,从而影响仓库运作效率。所以合理优化储位分配,提高AGV小车的作业效率是物流仓储中心的核心作业之一。本文首先介绍了储位分配问题研究的背景、意义及国内外储位分配优化的相关文献,并据此确定了本文的研究内容和研究方法。从优化储位分配模型及算法求解两个角度进行具体研究。以仓库内AGV搬运移动货架为研究背景,采用了SKU周转率原则与SKU关联性原则对货物进行储位分配。将高周转率货架存储到距离拣货口较近的位置上,将低周转率货架存储到距离拣货口较远的位置上;将相关性大、订购频次高的SKU就近存放在同一个货架上。最终构建了以SKU出入库距离最短、移动货架整体稳定性最高及相关SKU间距离最短的多目标储位分配优化模型。在求解储位分配优化模型时,首先根据仓库历史订单,采用FP-Growth算法计算SKU间的关联度和置信度。将频繁出现在同一订单的SKU存储在同一货架上。其次采用改进的多种群遗传算法对多目标储位分配优化模型求解。最后在实例分析部分,选取Y公司仓储中心为案例研究分析对象,结合实际的调研情况得到相关数据。采用遗传算法(GA)、多种群遗传算法(MPGA)和改进的多种群遗传算法(IMPGA)分别对本文储位分配优化模型一一求解。计算结果显示利用IMPGA算法对储位优化模型求解时,在SKU出入库距离、货架整体稳定性和相关SKU间距离三方面,IMPGA算法优化效果都更加显著。同时IMPGA算法避免了传统GA算法在求解问题时易出现早熟收敛等状况。因此在利用FP-Growth算法挖掘频繁项集的基础上,运用IMPGA算法进行储位优化,为现代化仓储中心的储位优化问题提供了新思路。更多还原

【Abstract】 With the rapid development of e-commerce,the degree of automation in warehouse centers get higher and higher.Aiming at the situation where the storage volume and variety of fast-moving goods in general large-scale warehouse centers are large,the use of AGV(Automated Guided Vehicle)trolleys the "person-to-person" picking operation model emerged at the historic moment and has been gradually adopted by many e-commerce storage centers.The use of AGV to carry mobile three-dimensional racks replaces the movement of staff.One rack can even be picked at the same time for multiple SKU,which improve the efficiency of order picking and reduce the cost of human resources.With the diversified changes in customer order requirements,the traditional single random storage location and classified storage location strategy can no longer meet the needs of storage location planning,which affects the efficiency of warehouse operations.Therefore,rationally optimizing the allocation of storage locations and improving the operation efficiency of AGV trolleys is one of the core operations of the logistics storage center.This article introduces the research background and significance of storage allocation and relevant literatures on storage allocation optimization,and then confirms the research content and research methods of this article.Specific research is carried out from two perspectives of optimized storage allocation model and algorithm solution.Taking the AGV handling mobile shelves in the warehouse as the research background,the SKU turnover principle and the SKU correlation principle are used to allocate the storage space of the goods.Store high turnover rate shelves closer to the picking port,and store low turnover rate shelves far away from the picking port;store SKUs with high relevance and high frequency of ordering as close as possible to the same place On the shelf.Finally,a multi-objective storage allocation optimization model was constructed with the shortest SKU entry and exit distance,the highest overall stability of the mobile shelf,and the shortest distance between related SKUs.When solving the storage allocation optimization model,first,according to the warehouse’s historical orders,the FP-Growth algorithm is used to calculate the correlation and confidence between SKUs,and the SKUs that frequently appear in the same order are stored on the same shelf.Secondly,an improved multi-population genetic algorithm is used to solve the multi-objective storage allocation optimization model.Finally,in the case analysis part,the Y company storage center is selected as the case study and analysis object,and the relevant data is obtained in combination with the actual research situation.Genetic algorithm(GA),multi-group genetic algorithm(MPGA)and improved multi-group genetic algorithm(IMPGA)are used to solve the optimal model of storage allocation in this paper one by one.The calculation results show that when the IMPGA algorithm is used to solve the storage location optimization model,the optimization effect of the IMPGA algorithm is more significant in terms of the SKU entry and exit distance,the overall shelf stability,and the distance between related SKUs.At the same time,IMPGA algorithm avoids the premature convergence of traditional GA algorithm when solving problems.Therefore,on the basis of using the FP-Growth algorithm to mine frequent it emsets,the IMPGA algorithm is used to optimize the storage location,which provides a new idea for the storage location optimization problem of modern storage centers.更多还原