A New Strategy to Improve Particle Swarm Optimization Exploration Ability
Sameh Kessentini, Dominique Barchiesi · 2010
To improve Particle swarm optimization (PSO) ability to explore new areas without delaying the algorithm convergence, a novel strategy is proposed which consists of choosing the best behavior while the new computed position of particle exceeds the search space. The strategy is tested and compared with conventional ones using adaptive PSO algorithm. Simulation results of benchmark functions are analyzed and show that the new strategy guarantees rapid exploration.