Izvorni znanstveni članak
Developing dynamic maximal covering location problem considering capacitated facilities and solving it using hill climbing and genetic algorithm
Jafar Bagherinejad
; University of Alzahra
Mehdi Seifbarghy
; University of Alzahra
Mahnaz Shoeib
; University of Alzahra
Sažetak
The maximal covering location problem maximizes the total number of demands served within a maximal service distance given a fixed number of facilities or budget constraints. Most research papers have considered this maximal covering location problem in only one period of time. In a dynamic version of maximal covering location problems, finding an optimal location of P facilities in T periods is the main concern. In this paper, by considering the constraints on the minimum or maximum number of facilities in each period and imposing the capacity constraint, a dynamic maximal covering location problem is developed and two related models (A, B) are proposed. Thirty sample problems are generated randomly for testing each model. In addition, Lingo 8.0 is used to find exact solutions, and heuristic and meta-heuristic approaches, such as hill climbing and genetic algorithms, are employed to solve the proposed models. Lingo is able to determine the solution in a reasonable time only for small-size problems. In both models, hill climbing has a good ability to find the objective bound. In model A, the genetic algorithm is superior to hill climbing in terms of computational time. In model B, compared to the genetic algorithm, hill climbing achieves better results in a shorter time.
Ključne riječi
maximal covering location problem; dynamic (multi-period) MCLP; capacitated MCLP; genetic algorithm; Hill climbing heuristic
Hrčak ID:
181512
URI
Datum izdavanja:
16.5.2017.
Posjeta: 1.364 *