Design of hybrid genetic algorithm with preferential local search for multiobjective optimization problems

Abstract

Evolutionary algorithms are used to obtain the optimal solutions for varieties of engineering problems. The performance of evolutionary algorithms can be enhanced by integrating them with local search methods. The idea is to combine the best of both global and local optimization approaches to perform a better exploration of search space. This paper presents a preferential hybrid evolutionary algorithm, where the gradient descent method is integrated into the NSGA-II. This new algorithm has been verified on a set of multiobjective benchmark problems using four different performance metrics. The results show that the proposed algorithm brings out the optimal solutions with better diversity and closeness to the known optimal solutions than NSGA-II and also consumes less time than traditional hybrid algorithm. © 2011 Springer-Verlag.