A bi-objective hub maximal covering location problem considering time-dependent reliability and the second type of coverage
A bi-objective hub maximal covering location problem considering time-dependent reliability and the second type of coverage
复制标题
考虑时变可靠性和第二类覆盖的双目标集线器最大覆盖位置问题
DOI：
10.1080/17509653.2015.1056265
复制
发表时间：
2016
影响因子：
4.8
通讯作者：
Mehri Sheikhi
中科院分区：
文献类型：
--
作者：
S. Pasandideh;S. T. A. Niaki;Mehri Sheikhi
摘要 集线器选址问题寻求集线器的最佳位置以及将非集线器节点分配给集线器。枢纽位置问题出现在各种应用中，包括航空系统、货物运输系统和电信网络设计。在本文中，考虑与时间相关的可靠性，提出了双目标枢纽最大覆盖位置问题。这两个目标函数是：（i）最大化加权网络可靠性和（ii）最大化枢纽网络中的总流量，其中用于定位的覆盖类型是第二类型。此外，每对节点之间的运输成本被假设为不确定参数。采用机会约束规划来制定双目标问题。使用目标达成方法（多目标决策过程中的一种技术）将该模型转换为单目标模型。由于该问题属于NP难问题，因此开发了遗传算法来解决它。由于文献中没有可用的基准，因此还开发了模拟退火算法以验证所获得的结果。利用响应面方法来调整两种算法的参数，以找到更好的解决方案。给出了一些数值例子来研究所提出算法的效率。最后，通过与理想解的相似度排序偏好技术对使用两种算法获得的结果进行比较。
Abstract The hub location problem seeks to find the best location for hubs and the assignment of non-hub nodes to hubs. The hub location problem appears in a variety of applications including airline systems, cargo delivery systems, and telecommunication network design. In this paper, a bi-objective hub maximal covering location problem is presented considering time-dependent reliabilities. The two objective functions are: (i) maximizing the weighted network reliability and (ii) maximizing the total flow in a hub network, where the type of coverage used for locating is the second type. Additionally, the transportation cost between each pair of nodes is assumed to be an uncertain parameter. Chance constrained programming is employed to formulate the bi-objective problem. The model is transformed into a single-objective model using the goal attainment method – a technique in multi-objective decision making procedures. As the problem belongs to the class of NP-hard problems, a genetic algorithm is developed to solve it. Since there is no benchmark available in the literature, a simulated annealing algorithm is developed as well in order to validate the results obtained. The response surface methodology is utilized to tune the parameters of both algorithms in order to find better solutions. Some numerical examples are presented to investigate the efficiency of the proposed algorithms. Finally, the results obtained using the two algorithms are compared by the technique for order preference by similarity to the ideal solution.