Hybridizing cultural algorithms and local search

Abstract

In this paper, we propose a new population-based framework for combining local search with global explorations to solve single-objective unconstrained numerical optimization problems. The idea is to use knowledge about local optima found during the search to a) locate promising regions in the search space and b) identify suitable step sizes to move from one optimum to others in each region. The search knowledge was maintained using a Cultural Algorithm-based structure, which is updated by behaviors of individuals and is used to actively guide the search. Some experiments have been carried out to evaluate the performance of the algorithm on well-known continuous problems. The test results show that the algorithm can get comparable or superior results to that of some current well-known unconstrained numerical optimization algorithms in certain classes of problems. © Springer-Verlag Berlin Heidelberg 2006.