A multi-objective particle swarm optimization algorithm with a dynamic hypercube archive, mutation and population competition

In this paper, an improved multi-objective particle swarm optimization algorithm (mPSO-DHA) with a dynamic hypercube archive (DHA), mutation and population competition is presented to enhance the performance of PSO in solving multi-objective optimization problems. The proposed algorithm considers a modification of the hyper-cube archiving method originally proposed in 2002 by Coello and Lechuga, and changes the bounds of the objective space dynamically in the optimization process. When the particles are trapped in local Pareto fronts, the algorithm introduces a mutation process in order to help the particles jump out. Also, weight adaptation and pool selection techniques are introduced in order to enhance the local searching ability. The proposed algorithm is applied to a series of wellknown benchmark problems, and results show that it can successfully find the true Pareto front with a good diversity of the solutions. In comparison to several other multi-objective particle swarm optimization algorithms, the proposed scheme showed better performance in solving benchmark functions.