A comparative study between CMA evolution strategies and Particle Swarm Optimization for antenna applications

Abstract

Nature-inspired optimization techniques have been at the forefront of research within electromagnetics due to their unique properties as global optimization algorithms. These algorithms are stochastic techniques which direct the optimizer towards the most likely position based on previously tested points. The biggest question for current researchers in this area is which algorithm performs the fastest, provides the best solution, and offers robust convergence for a variety of different function topologies. Within the domain of nature-inspired optimization techniques, the Covariance Matrix Adaptation (CMA) Evolution Strategies (ES) and the Particle Swarm Optimization (PSO) techniques have transpired due to their rapid convergence for many electromagnetics optimization problems.