A genetic algorithm with constrained sorting method for constrained optimization problems

Abstract

Engineering problems are commonly optimization problems with various constraints. For solving these constrained optimization problems, an effective genetic algorithm with a constrained sorting method is proposed in this work. The constrained sorting method is based on a dynamic penalty function and a non-dominated sorting technique that is used for ranking all the feasible and infeasible solutions in the whole evolutionary population. The proposed algorithm is tested on five well-known benchmark functions and three engineering problems. Experimental results and comparisons with previously reported results demonstrate the effectiveness, efficiency and robustness of the present algorithm for constrained optimization problems.