A hybrid population-based algorithm for solving the fuzzy capacitated maximal covering location problem. (March 2023)

• Record Type: Journal Article
• Authors: Aider, Méziane; Dey, Imene; Hifi, Mhand
• Is Part Of: Computers & industrial engineering. Volume 177(2023)
• Publication Date: 2023-03
• Subjects: Fuzzy -- Grey wolf -- Local search -- Location -- Optimization -- Population; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2023.108982
• Languages: English
• ISSNs: 0360-8352

Abstract

Abstract: In this paper, we study the fuzzy capacitated maximal covering location problem, where a hybrid population-based algorithm is proposed to solve it. An instance of the problem is represented by a set of customers in a given network with their distances and such that the coverage degree of facilities and the distance between customers are fuzzy. Its goal is to open (select) a subset of facilities to be located on customers such that the coverage of customers should be maximized. Two versions of a hybrid population-based algorithm are designed for tackling the aforementioned problem: (i) a modified grey wolf optimizer combined with exploration and exploitation strategies, and (ii) a modified grey wolf optimizer combined with a series of local searches and an upper bound delimiting wolf movements. Each version of the method enhances the quality of the solutions belonging to the current population by adding a series of local searches. Further, the exploration strategy is based upon the drop/rebuild operator for jumping from the current region to unvisited ones. Finally, the behavior of the proposed method is evaluated on a set of benchmark instances of the literature, where its provided results are compared to those reached by more recent methods of the literature and the state-of-the-art Cplex solver. Encouraging results have been obtained. Highlights: We study the fuzzy capacitated maximal covering location problem, an NP-hard problem. An efficient population-based algorithm is designed for tackling large-scale instances. A new hybridization is employed for highlighting the quality of the final solutions. An extensive experimental part is presented for evaluating the effectiveness of the resulting algorithm. The algorithm discovers several new bounds over the benchmark instances tested.