Vector Evaluated Particle Swarm Optimization Archive Management: Pareto Optimal Front Diversity Sensitivity Analysis

Vector evaluated particle swarm optimization (VEPSO) extends the particle swarm optimization (PSO) algorithm to deal with multi-objective optimization problems (MOPs). VEPSO stores the found non-dominated solutions in an archive. Management of the VEPSO archive is, however, not described in detail. In this paper, a Pareto optimal front (POF) diversity sensitivity analysis on the choice of the VEPSO's archive size and deletion approach is presented. The results indicate that the well-known spread, solution distribution, maximum spread, and spacing metrics are all sensitive to the choice of the archive size and deletion approach. It is concluded that care must be taken when selecting the archive size and deletion approach as the impact on the diversity of the POF is notable.