Electromagnetic Optimization Based on Gaussian Crow Search Approach
Leandro dos Santos Coelho, Carlos Eduardo Klein, Viviana Cocco Mariani, Carlos Augusto Richter do Nascimento, Alireza Askarzadeh · 2018
Metaheuristic algorithms have provided efficient tools to designers by which it became possible to determine the optimum solutions of optimization problems encountered in several fields such as engineering, finance and computation. Generally metaheuristics are based on metaphors that are taken from nature or some other processes. Recently, an optimization metaheuristic paradigm based on the social behavior of the crows, called crow search algorithm (CSA), has been proposed. The idea of CSA is motivated from the storing process of the excess food in hiding places then restoring it in the necessary time. The CSA has many favorable characteristics, such as ease of implementation, the need to adjust only a few parameters, a favorable balance between search diversification and intensification, and fast convergence and high sensitivity. However, to enhance the performance of the standard CSA approach, this paper proposes to tune the control parameters using population diversity information and normal (Gaussian) probability distribution function. Loney's solenoid benchmark and circular antenna array design problems are used to evaluate the effectiveness of the conventional CSA and the proposed variant. Simulation results and comparisons with the CSA demonstrated that the performance of the proposed MCSA approach is promising in electromagnetics optimization.