基于订单数据的库位优化研究

发布时间：2019-06-03 23:23

【摘要】：随着国民经济的快速发展,物流的重要性日益凸显。作为“十二五”期间国家重点扶持行业,中国物流业取得了快速的发展。统计数据显示,2013年全国社会物流总额达到197.8万亿元,物流业增加值达到3.9万亿元,物流业增加值占国内生产总值的比重由2005年的6.6%提高到2013年的6.8%,占服务业增加值的比重达到14.8%。物流业吸纳就业人数达到2890万人。中国物流业在过去十几年中取得的辉煌成绩不容忽视,但随着电子商务的快速发展,国内物流行业仍然存在基础设施不健全、管理水平低下等问题,2013年全社会物流总费用与国内生产总值的比率高达18%。互联网带来的新型商业模式给中国物流行业的发展提出了新的挑战,同时又为我们实现跨越式发展提供了机遇。本文以仓储管理中的库位优化问题为选题,根据某企业的订单数据进行库位优化建模分析。通过实验对不同的库位分配方案进行评价,并给出相关建议。本文的研究意义是探索仓储系统的库位分配方案,为企业的生产实践提供理论指导,提升国内仓储管理水平,进而降低仓储管理成本。本文有以下几个创新点：(1)同时使用三种货位分配方案对案例进行优化并将三种方案的优化效果对比,前人研究论文未曾有过；(2)对遗传算法进行改进,同时设计几种不同的交叉变异算子；(3)首次对随机存储这种模式进行了基于实际订单数据的模拟仿真；(4)在进行库位优化实证分析时首次将订单数据分为训练数据和验证数据。第一章介绍本文的研究背景,指出我国现阶段库位管理水平低下这一问题。针对这一问题提出本文的研究目的和意义,并对国内外研究现状进行综述。第二章详细介绍相关理论和方法。首先对现有库位优化理论进行分析,然后重点阐述ABC分类法和遗传算法两种理论。第三章我们对M配送中心的订单数据、现有的仓库布局以及库位分配方案进行分析,然后以订单拣选距离最小为目标,采用ABC分类思想对原有库位分配方案进行优化。通过实证分析,我们得出结论：新的分配方案优于原有的库位分配方案。第四章我们在原有案例背景的基础上,采用遗传算法进行库位分配。该部分我们为遗传算法设计了多种交叉变异算子进行仿真模拟,以保证算法的有效性和求解的全局性。通过实证分析,我们得出结论,遗传算法得到的库位分配方案优于原有的库位分配方案,且优于ABC库位分配方案。第五章我们基于订单数据流进行随机存储过程的模拟仿真,通过动态模拟订单出入库操作来检验随机存储模式的效率。仿真过程中我们设置不同的再订货点,结果表明随机存储优于原有的库位分配方案,且优于基于GA的库位分配方案。第六章我们对全文进行总结和展望。通过对比三种库位分配方案的实验结果,我们得出结论：(1)ABC分类法可以有效的缩短订单拣选距离,且使用简单便捷,方便实现；(2)基于改进遗传算法的货位分配方案可以有效缩短订单拣选距离,且优化效果明显；(3)随机存储模式效果最优,但技术硬件要求较高；(4)三种优化方案均能缩短订单拣选距离,提高操作效率,企业可根据仓库的实际情况选择库位分配方案。
[Abstract]:With the rapid development of national economy, the importance of logistics is becoming more and more prominent. As the key supporting industry in the "12th Five-Year" period, China's logistics industry has made rapid development. Statistics show that the total national social logistics in 2013 reached 197.8 trillion yuan, the added value of the logistics industry reached 3.9 trillion yuan, and the added value of the logistics industry accounted for 6.8% of the gross domestic product in 2005, accounting for 14.8% of the added value of the service industry. The number of employment in the logistics industry reached 280.9 million. China's logistics industry has achieved great achievements over the past decade, but with the rapid development of e-commerce, the domestic logistics industry still has the problems of inadequate infrastructure and low management level. The ratio of total social logistics cost to gross domestic product in 2013 is as high as 18 %. The new business model brought by the Internet poses a new challenge to the development of China's logistics industry, and also provides the machine for the development of the great-leap-forward development. In this paper, based on the stock-position optimization problem in the warehouse management, this paper is based on the order data of a certain enterprise to optimize the library position. The model analysis is carried out. The different library bit allocation schemes are evaluated by the experiment and the phase is given. The research significance of this paper is to explore the stock allocation scheme of the storage system, to provide the theoretical guidance for the production practice of the enterprise, to improve the management level of the storage in the country, and to further reduce the storage pipe. The paper has the following innovation points: (1) the case is optimized and the optimization effect of the three schemes is compared by using three kinds of locator allocation schemes, and the previous research papers have not been used; (2) the genetic algorithm is improved, and several different cross-changes are designed (3) For the first time, the model is simulated based on the actual order data. (4) The order data is divided into the training data and the test for the first time in the case of the case analysis. The first chapter introduces the background of this paper and points out that the management level of the library in our country is low at present. In this paper, the purpose and significance of this paper are put forward in this paper, and the research is made at home and abroad. a summary of the form. Chapter II Related theories and methods. First of all, the existing library bit optimization theory is analyzed, and then the ABC classification and the legacy are highlighted. In the third chapter, we analyze the order data, the existing warehouse layout and the location allocation scheme of the M distribution center, then take the minimum order picking distance as the target, and adopt the ABC classification thought to the original library. Through empirical analysis, we conclude that the new allocation scheme is superior to that of the new allocation scheme. The fourth chapter is based on the background of the original case. In this paper, a genetic algorithm is used to carry out the library bit allocation. The part of the algorithm is a genetic algorithm, and a variety of cross-mutation operators are designed for simulation, so as to ensure the algorithm. Based on the empirical analysis, we draw a conclusion that the base allocation scheme obtained by the genetic algorithm is superior to that of the original library allocation scheme, and The fifth chapter is based on the simulation of the order data flow and the simulation of the random storage process. In the process of simulation, we set different re-order points, and the results show that the random storage is better than the original location allocation scheme. And is superior to a GA-based library bit allocation scheme. In chapter 6, we sum up and view the full text. By comparing the experimental results of the three kinds of location allocation schemes, we can conclude that: (1) ABC classification can effectively shorten the order picking distance, and it is simple and convenient to implement; and (2) based on the improved genetic algorithm. The scheme can effectively shorten the order picking distance, and the optimization effect is obvious; (3) the random storage mode effect is optimal, but the technical hardware requirement is high; (4) the three optimization schemes can shorten the order picking distance and improve the operation efficiency; and the enterprise can
【学位授予单位】：浙江工商大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F259.2

【参考文献】