The Self-adaptive Comprehensive Learning Particle Swarm Optimizer.
Adiel Ismail, Andries Petrus Engelbrecht · 2012
Particle swarm optimization (PSO) has been applied successfully to a wide range of optimization problems. Appropriate values for control parameters of the particle swarm optimization (PSO) algorithm are critical to its success. This paper proposes that the control parameters of PSO be embedded in the position vector of particles and dynamically adapted while the search is in progress, relieving the user from specifying appropriate values before the search commences. Application of the Self-Adaptive Comprehensive Learning Particle Swarm Optimizer (SACLPSO) to 9 well known test functions show an improvement in performance on most of the functions compared to CLPSO and a tuned PSO.