A decoder-based evolutionary algorithm for constrained parameter optimization problems

Abstract

Several methods have been proposed for handling nonlinear constraints by evolutionary algorithms for numerical optimization problems; a survey paper [7] provides an overview of various techniques and some experimental results, as well as proposes a set of eleven test problems. Recently a new, decoder-based approach for solving constrained numerical optimization problems was proposed [2, 3]. The proposed method defines a homomorphous mapping between n-dimensional cube and a feasible search space. In [3] we have demonstrated the power of this new approach on several test cases. However, it is possible to enhance the performance of the system even further by introducing additional concepts of (1) nonlinear mappings with an adaptive parameter, and (2) adaptive location of the reference point of the mapping.