Dynamic Population Strategy Assisted Particle Swarm Optimization in Multiobjective Evolutionary Algorithm Design

In this research report, the author proposes two new evolutionary approaches to Multiobjective Optimization Problems (MOPs)— Dynamic Particle Swarm Optimization (DPSMO) and Dynamic Particle Swarm Evolutionary Algorithm (DPSEA). In DPSMO, instead of using genetic operators (e.g., crossover and mutation), the information sharing technique in Particle Swarm Optimization is applied to inform the entire population more accurate moving direction and speed than EA. Meanwhile, based on the dynamic population strategies, cell-based rank and density estimation and objective space compression strategy used in Dynamic Mutiobjective Evolutionary Algorithm (DMOEA), the DPSMO can also evolve to an approximately optimal population size while the population is approaching the true Pareto front. Moreover, to overcome DPSMO’s difficulty in producing highquality Pareto front, DPSEA is designed by combining both EA and PSO’s information sharing techniques. By examining the selected performance measures on two test functions, DPSEA is found to be competitive with, or even superior to DMOEA and DPSMO in terms of keeping the diversity of the individuals along the trade-off surface, tending to extend the Pareto front to new areas and finding a well-approximated Pareto optimal front. Moreover, from the experimental results, DPSMO and DPSEA also show more promising performance in improving algorithm efficiency comparing to DMOEA.