Dynamic Particle Swarm Optimization Based on Anti-Mistake and Sinai Chaos

To solve the premature problem of particle swarm optimization, firstly, the dynamic nonlinear inertia weights are designed which can make particles retain the favorable conditions and converge to the global optima continually; secondly, two kinds of anti-mistake equations are introduced which can make the stagnated particles break away from the local optima and dynamically search the global optima; at the same time, the Sinai chaos is introduced which can enrich search behaviors and make particles travel the whole search space. Experimental results demonstrated that the new introduced methods outperformed several other improved particle swarm optimization algorithms on many famous benchmark problems.