Um Estudo Sobre Meta-Heurísticas em Otimização Multimodal

Abstract

RESUMOThis paper presents an investigation of four metaheuristics for multimodaloptimization. The algorithms have been implemented inR and compared against each other using well-known benchmarkfunctions. We implemented the following algorithms: Genetic Algorithm(GA), Particle Swarm Optimization (PSO), Differential Evolution(DE), and Bat Algorithm (BA). For comparisons, we used fivemultimodal benchmarks: Rosenbrock, Griewank, Ackley, Schwefel,and Alpine. Preliminary results show that Bat Algorithm and GeneticAlgorithm tend to discover the best solution considering theparameters that have been set up.