Probability Diffused Particle Swarm Optimization

摘要

Premature convergence is a major problem of Particle Swarm Optimization (PSO).Although many strategies have been proposed, there is still some work needed to do in high-dimensional cases. To overcome this shortcoming, a diffused velocity update equation is designed aiming to improve the population diversity with a probability threshold. Simulation results show the performance of this new variant is superior to other two previous modifications when dealing with multi-modal high-dimensional benchmark functions.