Skip to Main Content
This paper proposes a hybrid multiobjective particle swarm approach called fuzzy clustering multi-objective particle swarm optimizer (FC-MOPSO). This model uses a fuzzy clustering technique in order to provide a better distribution of solutions in decision variable space by dividing the whole swarm into subswarms. Furthermore, fuzzy clustering technique offers a natural way to deal with overlapping clusters and does not require prior information on data distribution. Each sub-swarm has its own set of leaders and evolves using the PSO algorithm and the concept of Pareto dominance. In FC-MOPSO, the migration concept is performed in order to exchange information between different subswarms and ensure their diversity. The proposed algorithm is compared with other multiobjective particle swarm optimization algorithms on tree test functions. The results show that the proposed algorithm attains better performance of convergence and diversity.