自动化立体库货位分配与Flexsim仿真研究

发布时间：2018-04-05 04:09

本文选题：货位分配　切入点：自动化立体库　出处：《吉林大学》2017年硕士论文

【摘要】：自动化立体库存储货物数量庞大、品种繁多,事先不对货位进行规划,势必导致仓库作业的拥堵、中断,进而影响整个仓库作业效率。随着市场需求和企业的生产经营活动的变化,自动化立体库中的货物在数量和种类上也会发生相应的变化,在时间累积效应下,原有的货位分配方案不再适用,需要对货位进行重新分配。因此,对货物进行货位分配是自动化立体库管理活动中重要的决策内容之一。目前绝大多数研究学者在研究货位分配问题时,基于货架整体的稳定性、出入库效率、分巷道存储以及货物关联性等货位分配原则进行考虑,只以货架整体的稳定性和货物出入库效率为目标,建立相应的货位分配数学模型,然而,并没有过多地在数学模型中考虑其他分配原则对货位分配的影响。因此,本文在提高货架整体稳定性和出入库效率目标的基础上,同时考虑了货物之间的关联性,即在进行货位分配时同时要求具有关联性的货物在货架上就近存放。在求解货位分配数学模型时,多采用了简单加权遗传算法,这种方法忽视了各目标函数之间的单位不统一的情况,同时也容易陷入局部最优,出现未成熟收敛问题。针对这些问题,本文在归一化处理多个货位分配优化目标的同时,引入多种群遗传算法求解,解决了各目标函数之间单位不统一,并避免了遗传算法未成熟收敛问题。本文对自动化立体库货位分配问题的研究现状进行了综述,分析了货位分配对仓库作业的影响,对自动化立体库构造、货物存储模式和货位分配的相关原则归纳总结,提出了本文货位优化分配的三个目标:提高货物出入库效率、降低货架整体的等效重心和关联性货物就近存储;针对提出的货位分配优化目标,建立了相应的多目标货位分配数学模型,并详细设计多种群遗传算法进行求解;结合魔幻战蝠车模流水线装配实验的零件数量和重量信息,对自动化立体库的货位分配数学模型求解,分别得到多种群遗传算法和简单加权遗传算法两组求解方案,并对货位分配方案对应的目标函数值和算法效率进行了对比分析,验证了多种群遗传算法在求解结果准确性和稳定性的优越性。最后,结合Flexsim软件对算例货位分配方案进行模拟,从动态的角度验证了货位分配方案的有效性。
[Abstract]:The quantity and variety of the goods stored in the automated warehouse are huge, so it will inevitably lead to the congestion and interruption of the warehouse operation, which will affect the efficiency of the whole warehouse operation.With the change of market demand and enterprise's production and management activities, the quantity and type of goods in the automated warehouse will also change correspondingly. Under the effect of time accumulation, the original distribution scheme of goods is no longer applicable.The cargo space needs to be reallocated.Therefore, the distribution of goods is one of the important decision-making contents in automated warehouse management.At present, most of the research scholars in the study of the allocation of cargo space, based on the overall shelf stability, storage efficiency, sub-laneway storage and cargo correlation and other cargo allocation principles are considered.Aiming at the stability of the whole shelf and the efficiency of goods entering and leaving, the mathematical model of the distribution of goods space is established. However, the influence of other allocation principles on the allocation of goods is not considered too much in the mathematical model.Therefore, on the basis of improving the overall stability of shelves and the efficiency of entering and leaving storage, this paper also takes into account the correlation between goods, that is, the goods with relevance are required to be stored nearby on the shelves during the distribution of goods.A simple weighted genetic algorithm is used to solve the mathematical model of cargo location allocation. This method ignores the unity of the units among the objective functions and is prone to fall into local optimum and appear immature convergence problem at the same time.In order to solve these problems, in order to solve the problem of unit disunity among objective functions and avoid the premature convergence of genetic algorithm, this paper normalizes the optimization target of multiple cargo allocation, and introduces multi-population genetic algorithm to solve the problem of multi-population genetic algorithm.In this paper, the current situation of research on the distribution of freight spaces in automated storage is summarized, and the influence of space allocation on warehouse operations is analyzed, and the construction, storage mode and relevant principles of distribution of goods are summarized.In this paper, three objectives of optimal allocation of goods are put forward: to improve the efficiency of loading and unloading, to reduce the equivalent center of gravity of the whole shelf and to store the relevant goods nearby, and to optimize the allocation of goods,The corresponding mathematical model of multi-target cargo space allocation is established, and the multi-population genetic algorithm is designed to solve the problem in detail, combining with the information of the number and weight of the parts in the assembly experiment of pipeline, the model of magic bat.To solve the mathematical model of freight location assignment of automated warehouse, two groups of solutions are obtained, namely, multi-population genetic algorithm and simple weighted genetic algorithm, and the objective function value and the efficiency of the algorithm are compared and analyzed.The superiority of multi-population genetic algorithm in the accuracy and stability of the solution is verified.Finally, the effectiveness of the allocation scheme is verified from the dynamic point of view by simulating the distribution scheme with Flexsim software.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274;TP18

【相似文献】