Global Optimization by Particle Swarm Method
Sudhanshu Mishra · SSRN Electronic Journal · 2010
This paper introduces the problem of global optimization of the complicated multimodal nonlinear objective functions by stochastic population-based methods, such as Genetic algorithms, classical and generalized simulated annealing (GSA), and differential evolution (DE) in general, and the particle swarm (PS) method in particular. It illustrates how an extremely nonlinear Chacon-Gielis function involving deeply nested Jacobian elliptic integrals can be fitted to empirical data by the repulsive particle swarm (RPS) method.