An evolutionary algorithm for constrained Bi-objective optimization using radial slots

Abstract

In this paper, we introduce an evolutionary algorithm for constrained, bi-objective optimization. The objective space is divided into a predefined number of radial slots and solutions compete with members in the same slot for existence. The procedure creates a uniform spread of solutions across the slots and they collectively form the nondominated front. Constraints are handled using a standard min-max formulation. We report the performace of our algorithm on a set of seven constrained, bi-objective test problems (CTP1 to CTP7) which have been known to pose difficulties to all existing multiobjective algorithms. © Springer-Verlag Berlin Heidelberg 2005.