Simplified tabu search with random-based searches for bound constrained global optimization

Abstract

This paper proposes a simplified version of the tabu search algorithm that solely uses randomly generated direction vectors in the exploration and intensification search procedures, in order to define a set of trial points while searching in the neighborhood of a given point. In the diversification procedure, points that are inside any already visited region with a relative small visited frequency may be accepted, apart from those that are outside the visited regions. The produced numerical results show the robustness of the proposed method. Its efficiency when compared to other known metaheuristics available in the literature is encouraging.