Efficient simulated annealing based solution approaches to the competitive single and multiple allocation hub location problems

Abstract

Hub location problems (HLPs) constitute an important class of problems in logistics with numerous applications in passenger/cargo transportation, postal services, telecommunications, etc. This paper addresses the competitive single and multiple allocation HLPs where the market is assumed to be a duopoly. Two firms (decision makers) sequentially decide on the configuration of their hub networks trying to maximize their own market shares. The customers choose one firm based on the cost of service provided by these firms. Mathematical formulations are presented for the problems of the first and second firms (the leader and the follower, respectively) and Simulated Annealing (SA) based solution algorithms are proposed for solving these problems both in single and multiple allocation settings. Extensive computational experiments show the capability of the proposed solution algorithms to obtain the optimal solutions in short computational times. Some managerial insights are also derived based on the obtained results.

A Tabu Search Based Solution Approach to the Competitive Multiple Allocation Hub Location Problem

“…Since CPLEX was not able to solve any of the instances within the time limit of 5 hours, no result is not reported for its performance. The results in Table 4 show that the proposed TS algorithm outperforms the SA [10] algorithm both in terms of solution time and quality. It can be seen that for seven instances (denoted in bold in Table 4), the objective values of the solutions found by the TS are better than those of the SA's solution.…”

“…However, due to the large size of these instances (as it now includes all the 81 nodes of the TR data set as candidate locations for installing hubs), CPLEX was not able to solve any of the instances within the time limit of 5 hours. Therefore, to evaluate the efficiency of the proposed TS algorithm, we solved the problem using an existing simulated annealing (SA) based algorithm proposed in [10]. The results for solving the problem using the TS and SA algorithms with the TR data set by assuming | | = 81 are reported in Table 4.…”

“…Both problems were formulated as MIP models and exact solution algorithms based on enumeration were proposed for solving them. More recently, Ghaffarinasab et al [10] addressed the competitive single and multiple allocation HLPs as Stackelberg games and proposed efficient simulated annealing (SA) based heuristics for solving the problems in both single level (follower's problem) and bilevel (leader's problem) settings.…”

“…Constraints (9) state that the flows can only be routed via nodes that have been designated as hubs. The constraints (10) and (11) define the real and binary variables, respectively.…”

Strategic territorial deployment of hospital pharmacy robots using a stochastic p-robust optimization approach

“…The problem consists in determine the optimal policy that indicates the sources and destinations of transshipments where the demand is stochastic. Other applications for different variants of location-allocation problems can be found in [23]- [26].…”