RSM enhanced ACO R algorithm for electromagnetic optimization

This paper studies and proposes the use of RSM (Response Surface Methodology) to enhance the performances of the ACO R (Ant Colony Optimization for Real domains) algorithm. Several RSM models, such as CCD (Central Composite Design), Box-Behnken and Doelhert, are employed and applied to construct a metaheuristic function in order to find the optimal set of parameters for an ACO R algorithm when solving an electromagnetic problem. The problem used for testing is a benchmark proposed by the electromagnetic community namely the Loney’s solenoid. The ACO R optimal parameters are searched for two different objectives, the best solution and the average best solution over a set of tests.