《Handling constraints in global optimization using and AIS》.pdf

Handling Constraints in Global Optimization using an Artificial Immune System Nareli Cruz-Cortes,´ Daniel Trejo-Perez´ and Carlos A. Coello Coello CINVESTAV-IPN, Evolutionary Computation Group Depto. de Ingenierıa´ Electrica,´ Seccion´ de Computacion´ Av. Instituto Politecnico´ Nacional No. 2508 Col. San Pedro Zacatenco, Mexico,´ D. F. 07360, MEXICO nareli,dtrejo @computacion.cs.cinvestav.mx,
ccoello@cs.cinvestav.mx Abstract. In this paper, we present a study of the use of an artificial immune system (CLONALG) for solving constrained global optimization problems. As part of this study, we evaluate the performance of the algorithm both with binary encoding and with real-numbers encoding. Additionally, we also evaluate the im- pact of the mutation operator in the performance of the approach by comparing Cauchy and Gaussian mutations. Finally, we propose a new mutation operator which significantly improves the performance of CLONALG in constrained op- timization. 1 Introduction Many bio-inspired algorithms (particularly evolutionary algorithms) have been very successful in the solution of a wide variety of optimization problems [3]. However, all of these approaches (including evolutionary algorithms and artificial immune systems), when used for numerical optimization, can be seen as unconstrained search techniques. This means that they require a suitable mechanism to incorporate constraints, such that they can deal with the general nonlinear optimization problem. Within evolutionary algorithms (EAs), external penalty functions have been the most popular mechanism adopted to incorporate constraints into the fitness function [17]. The idea of an external p