Comparison of local search-based metaheuristics on the multiple objective knapsack problem

The paper presents a comparative experiment with local search-based metaheuristics on the multiple objective knapsack problem. The methods used in this experiment are: Memetic-Pareto Archive Evolution Strategy (M-PAES), multiple objective genetic local search (MOGLS), Serafmi's multiple objective simulated annealing (SMOSA), multiple objective simulated annealing proposed by Ulungu et al. (MOSA) and Pareto simulated annealing (PSA). The experiment follows the design of the previous comparative studies in which M-PAES proved to be the best performer among several evolutionary algorithms. The results of the experiment reported in this paper indicate the best performance of MOGLS and PSA on the multiple objective knapsack problem.