基于改进蚁群算法的物流网络

摘要

文中将受容量限制的单分配轴-辐式网络抽象为一个三次变量的混合整数线性规划模型方程；提出了一种改进的蚁群算法，将6种局域搜索算子加入算法中，因此具有较高的全局搜索能力和局部搜索能力；同时提出“解对”的概念，对问题的构成进行分解优化，转化为确定问题，切实使本问题符合蚁群算法使用的前提和优势；最后，使用澳大利亚邮政的数据进行选址仿真实验，验证此算法模型在该应用中的求解效率和计算稳定性。%Capacitated single allocation hub-and-spoke networks can be abstracted as a mixed integer linear pro-gramming model equation with three variables.By introducing an improved ant colony algorithm which has six lo-cal search operators and the "Solution Pair"concept to decompose and optimize the composition of the problem, it can become specific and more effective to meet the premise and advantages of using ant colony algorithm.Fi-nally,location simulation experiment is made with Australia Post data to demonstrate that this algorithm has high efficiency and stability for solving this problem.