Dynamic Search With Charged Swarms
Tim Blackwell, Peter J. Bentley · 2002
Two novel particle swarm optimization (PSO) algorithms are used to track and optimize a 3-dimensional parabolic benchmark function where the optimum location changes randomly and with high severity. The new algorithms are based on an analogy of electrostatic energy with charged particles. For comparison, the same experiment is performed with a conventional PSO algorithm. It is found that the best strategy for this particular problem involves a combination of neutral and charged particles. 1