By Rocha, AMAC; Fernandes, EMGP; Revetria, R; Cecchi, A; Schenone, M; Mladenov V; Zemliak, A
Proceedings of the 6th Wseas International Conference on System Science and Simulation in Engineering (Icossse '07)
The Electromagnetism-like (EM) algorithm, developed by Birbil and Fang [31 is a population-based stochastic global optimization algorithm that uses an attraction-repulsion mechanism to move sample points toward optimality. In order to improve the accuracy of the solutions the EM algorithm incorporates a random local search. In this paper we propose: a new local search procedure based on a pattern search method, and a population shrinking strategy to improve efficiency. The proposed method is applied to some test problems and compared with the original EM algorithm.