Output details

A novel selection evolutionary strategy for constrained optimization

This paper presents a novel evolutionary algorithm strategy for constrained optimisation problems (COPs). Typically COPs are solved with algorithms that, firstly, search for a “feasible” solution (one which satisfies the constraints), and secondly perform optimisation within the feasible region of the problem space. The novelty of this work lies in showing that infeasible solutions may still provide useful information to speed up optimisation within the feasible region. Evolutionary algorithms allow us to maintain a population containing both feasible and infeasible solutions, and combine both of these kinds of information to significantly outperform other state of the art methods.