多枢纽通道轴辐式物流网络设计研究

发布时间：2019-02-17 09:58

【摘要】：“如何降低物流费用,提高物流系统运作效率”这一问题早已不再只停留于理论研究层面转而以一种更加迅速且直接的方式影响着人们的日常生活。“互联网+”时代的到来、电子商务的快速崛起,“网购”与“虚拟经济”这些名词已经成为人们的一种消费与生活依赖。物流作为中间桥梁与纽带,其效率与成本从未如此深刻的影响过我们。而随着这一生活习惯的改变、伴随世界贸易大发展的背景,带来的同时还有对物流网络这一基础保障的巨大挑战。我国政府开始提倡“物流大通道”的宏观建设战略,而相关企业则开始新一轮的网络布局微观重构,两者都期待自上而下的打通物流网络大通道,希望从根源上降低社会物流成本、提升物流效率。在这一历史背景下,考虑我国社会物流的发展现状,本文研究了一类特殊的轴辐式物流网络设计问题——多枢纽通道轴辐式物流网络。在该问题中,由于我国现阶段阶段性投资限制及国际物流网络无法克服的地理因素制约的事实存在,网络中枢纽节点间无法形成完全联通的枢纽级别网络。因此,本文考虑枢纽节点的选址必须形成一系列相互连接的通道网络,通过连接枢纽间高效的通道实现运输成本或运输时间的规模经济效益。提出了一个纯整数非线性规划模型用来求解网络设计中的枢纽选址及通道配置问题。在考虑所有枢纽网络关键节点处的物流作业时间消耗前提下,研究的目标是最小化所有节点间总的加权运输时间,同时考虑物流枢纽网络建设总成本的预算约束,使其更加符合我国当前的物流网络设计现状与需求。本文提出了混合精英遗传算法(Hybrid Elitist Genetic Algorithm,HEGA)用以求解上述模型。该算法融合传统遗传算法的基本思想并引入精英保留策略,将模拟退火算法中的Metropolis接受准则及粒子群算法中的更新规则引入遗传操作,得到了一种兼具全局搜索能力、有效收敛性能、精确且稳定的启发式算法。通过比较LINGO与HEGA算法同等规模计算实验的结果,得到了一致的计算结果以及HEGA算法稳定理想的收敛曲线,证明了本文所提出的HEGA算法解决小规模问题的求解精度与收敛性,并展现了其超越精确算法的求解速度。中小规模及大规模标准算例的计算实验则验证了HEGA算法的求解效果,无论从收敛性、精确性还是稳定性上HEGA算法在解决本文所提出的多枢纽通道轴辐式物流网络设计问题中均有更为优秀的表现。本文旨在通过以上研究提出一类符合时代发展需求的物流网络规划与设计方法。
[Abstract]:The problem of "how to reduce the cost of logistics and improve the efficiency of logistics system" is no longer confined to the theoretical research level, but affects people's daily life in a more rapid and direct way. With the advent of the "Internet" era, the rapid rise of e-commerce, the terms "online purchase" and "virtual economy" have become a kind of consumption and life dependence. Logistics as an intermediate bridge and link, its efficiency and cost have never so profoundly affected us. With the change of living habits and the background of the great development of world trade, there is also a huge challenge to the basic guarantee of logistics network. Our government began to advocate the macro-construction strategy of "big logistics channel", while the related enterprises began a new round of micro-reconstruction of network layout. Both of them look forward to opening up the large logistics network channels from top to bottom. I hope to reduce the cost of social logistics and improve the efficiency of logistics. Under this historical background, considering the present situation of the development of social logistics in China, this paper studies a special kind of axle-spoke logistics network design problem, that is, multi-hub channel axis-spoke logistics network. In this problem, due to the limitation of investment and the fact that the international logistics network can not be overcome by geographical factors at present, the hub level network with complete connectivity can not be formed between the hub nodes in the network. Therefore, in this paper, a series of interconnected channel networks must be formed for the location of hub nodes, which can realize the economies of scale of transportation cost or transport time by connecting efficient channels between hubs. A pure integer nonlinear programming model is proposed to solve the problem of hub location and channel configuration in network design. The objective of the study is to minimize the total weighted transportation time between all the nodes and to consider the budget constraints of the total cost of the construction of the logistics hub network, taking into account the time consumption of logistics operations at the key nodes of the hub network, and the objective of the study is to minimize the total weighted transportation time between all the nodes. Make it more in line with our current logistics network design status and demand. In this paper, a hybrid elite genetic algorithm (Hybrid Elitist Genetic Algorithm,HEGA) is proposed to solve the above model. This algorithm combines the basic idea of traditional genetic algorithm and elite reservation strategy, and introduces the Metropolis acceptance criterion and the update rule of particle swarm optimization algorithm into genetic operation, and obtains a global search ability. Efficient convergence, accurate and stable heuristic algorithm. By comparing the experimental results of LINGO and HEGA algorithm in the same scale calculation, the consistent calculation results and the stable and ideal convergence curve of HEGA algorithm are obtained. The accuracy and convergence of the HEGA algorithm proposed in this paper for solving small scale problems are proved. It also shows the speed of solving the problem beyond the exact algorithm. The calculation results of small and medium scale and large scale standard examples verify the effectiveness of the HEGA algorithm, regardless of its convergence. In terms of accuracy and stability, the HEGA algorithm has a better performance in solving the multi-hub channel axis-spoke logistics network design problem proposed in this paper. The purpose of this paper is to put forward a class of logistics network planning and design methods that meet the needs of the times.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F259.2;U115

【参考文献】