A boundary search based ACO algorithm coupled with stochastic ranking

In this paper we present a boundary search based ACO algorithm for solving nonlinear constrained optimization problems. The aim of this work is twofold. Firstly, we present a modified search engine which implements a boundary search approach based on a recently proposed ACO metaheuristic for continues problems. Secondly, we propose the incorporation of the stochastic ranking technique to deal with feasible and infeasible solutions during the search which focuses on the boundary region. In our experimental study we compare the overall performance of the proposed ACO algorithm by including two different complementary constraint-handling techniques: a penalty function and stochastic ranking. In addition, we include in our comparison of results the stochastic ranking algorithm, which was originally implemented using an evolution strategy as its search engine.