基于改进的引力搜索算法求解物流配送中心选址问题

发布时间：2018-10-09 08:46

【摘要】：蓬勃发展的现代物流业,已经成为衡量国家经济实力的重要组成部分。物流配送中心在整个供应链中充当着重要的角色,把供应链上游的供应商与下游的零售商(或分销商)连成一个有机的系统。恰当的物流配送中心选址方案不仅能降低其自身的运营费用,还能够提升配送效率,增强竞争水平和发展潜力。此外,选址方案对与物流配送中心有业务往来的供应商、分销商和零售商的物流成本费用有很大的影响。河南省政府响应国家对物流基础设施建设的要求,制定并实施洛阳、商丘、南阳、信阳、安阳、濮阳、三门峡7个物流节点城市建设方案,建成投用15个区域分拨和城市物流配送中心。所以,对物流配送中心选址进行研究可以促进整个物流网络的协调发展。引力搜索算法是由E.Rashedi教授受牛顿万有引力定律和运动学第二定律的启发,于2009年提出的一种新的智能优化算法。该算法基于万有引力定律和粒子间的相互吸引现象,认为空间中的每个粒子都受到其他粒子的引力作用,且使空间中的粒子均朝引力最大的粒子方向移动。随着不断地进行移动,粒子最终聚集于质量最大的粒子附近,质量最大的粒子的位置即为所求问题的最优解。研究发现,在进行标准测试函数优化试验中,引力搜索算法的寻优精度和收敛速度均显著优于遗传算法和粒子群算法。本文的研究目的是把改进的引力搜索算法应用到三个不同的物流配送中心选址问题中,并将运算结果与标准引力搜索算法和粒子群算法进行对比,并为此开展了以下几项工作:(1)首先介绍了本文的选题背景和意义,以及物流配送中心选址问题国内外的研究现状。物流配送中心的定义、类型和功能以及物流配送中心选址的基本原则和方法。为接下来的研究内容做好理论铺垫。(2)针对引力搜索算法收敛速度慢、收敛精度低的问题。本章引入距离阈值的概念,衡量单个粒子与当代最优粒子间的距离,每个粒子独立动态地调整引力系数,实现了算法全局搜索能力和局部搜索能力的动态平衡,提高了算法的收敛速度。针对算法易早熟收敛的问题,提出了一种基于粒子聚集程度的变异策略,该策略通过参数α的取值有效避免了算法陷入早期收敛的问题。最后,使用基准测试函数实验验证了改进引力搜索算法在算法收敛速度和收敛精度方面取得的显著提高。(3)最后将改进的引力搜索算法运用于求解单一设施、连续选址问题;多设施、离散选址问题;多设施、连续、有约束选址问题。编写求解模型的改进引力搜索算法MATLAB程序,并通过对具体的选址案例的求解结果与标准引力搜索算法结果和粒子群算法结果对比,验证改进引力搜索算法求解此类问题的可行性,为求解物流配送中心选址问题提供一种新的求解方法。
[Abstract]:The booming modern logistics industry has become an important part of measuring the national economic strength. Logistics distribution center plays an important role in the whole supply chain, connecting the upstream suppliers of the supply chain with the downstream retailers (or distributors) into an organic system. Proper location of logistics distribution center can not only reduce its own operating costs, but also enhance the efficiency of distribution, enhance the level of competition and development potential. In addition, the location scheme has a great impact on the logistics cost of suppliers, distributors and retailers who do business with the logistics distribution center. In response to the requirements of the state for the construction of logistics infrastructure, the Henan Provincial Government has formulated and implemented seven urban construction schemes for logistics nodes in Luoyang, Shangqiu, Nanyang, Xinyang, Anyang, Puyang and Sanmenxia. Built 15 regional distribution and urban logistics distribution center. Therefore, the study of logistics distribution center location can promote the coordinated development of the whole logistics network. The gravitational search algorithm is a new intelligent optimization algorithm proposed by Professor E.Rashedi in 2009 inspired by Newton's law of gravitation and the second law of kinematics. Based on the law of universal gravitation and the phenomenon of mutual attraction between particles, the algorithm considers that each particle in space is subjected to the gravitational action of other particles, and makes the particles in space move towards the direction of the most attractive particle. With the continuous movement, the particles finally gather near the particle with the largest mass, and the position of the particle with the largest mass is the optimal solution to the problem. It is found that the optimization accuracy and convergence speed of the gravitational search algorithm are significantly better than those of the genetic algorithm and the particle swarm optimization algorithm. The purpose of this paper is to apply the improved gravity search algorithm to three different logistics distribution center location problems, and compare the results with standard gravitational search algorithm and particle swarm optimization algorithm. The following works are carried out: (1) the background and significance of this paper and the research status of logistics distribution center location at home and abroad are introduced. The definition, type and function of logistics distribution center and the basic principles and methods of location of logistics distribution center. It lays the foundation for the following research content. (2) aiming at the problem of slow convergence speed and low convergence precision of gravity search algorithm. In this chapter, the concept of distance threshold is introduced to measure the distance between a single particle and the contemporary optimal particle. Each particle adjusts the gravitational coefficient independently and dynamically, and realizes the dynamic balance between the global search ability and the local search ability of the algorithm. The convergence rate of the algorithm is improved. Aiming at the problem of premature convergence of the algorithm, a mutation strategy based on particle aggregation degree is proposed, which effectively avoids the problem of the algorithm falling into early convergence through the parameter 伪. Finally, the benchmark function experiment is used to verify the remarkable improvement of convergence speed and convergence accuracy of the improved gravitational search algorithm. (3) finally, the improved gravitational search algorithm is applied to solve the single facility and continuous location problem. Multi-facility, discrete location problem, multi-facility, continuous, constrained location problem. The improved gravitational search algorithm (MATLAB) program is written to solve the model, and the results are compared with the results of standard gravitational search algorithm and particle swarm optimization algorithm. It is proved that the improved gravitational search algorithm is feasible to solve this kind of problem, which provides a new method for solving the location problem of logistics distribution center.
【学位授予单位】：河南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F259.2

【参考文献】