Abstract

AbstractOptimization problems with many objectives open new issues for multi-objective optimization algorithms and particularly Particle Swarm Optimization. Many of the existing algorithms are able to solve problems of low number of objectives, but as soon as we increase the number of objectives, their performances get even worse than random search methods. This paper gives an overview on Multi-objective Particle Swarm Optimization when having many objectives and parameters. Furthermore, two new variants of MOPSO are proposed which are based on ranking of the non-dominated solutions. The proposed distance based ranking in MOPSO improves the quality of the solutions for even very large objective and parameter spaces. The quality of the new proposed MOPSO methods has been tested and compared to the random search and NSGA-II methods. The tests cover 3 to 20 objectives and 20 to 100 parameters.KeywordsParticle Swarm OptimizationMultiobjective OptimizationRanking MethodTournament SelectionRandom Search MethodThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call