A Multi-swarm Based Hybird Optimization Algorithm in Dynamic Environments

This paper analyzes the effect of population size to PSO algorithm and proposes Dynamic particle population based particle swarm optimization (DPP-PSO),the core idea of which is that, according to the search, particle swarm dynamically change particle population and gradually decrease the particles with lower search ability when population size are converging constantly and gradually increase new particles to expand the global search ability. The collaborative optimization of these items decrease calculated amount and improve globe exploration ability, which was proved in the test on four benchmark functions.