An improved Shuffled Frog Leaping Algorithm with a fast search strategy for optimization problems

Several evolutionary algorithms (EAs) are proposed in the literature to solve continuous optimisation problems. In this paper we present a new search startegy to improve the efficiency of the Shuffled Frog Leaping Algorithm (SFLA). The shuffled frog leaping algorithm is a population-based approach for a heuristic search in optimization problems. The algorithm consists of a set of virtual frogs partitioned into several groups called “memeplexes”. However, After some optimization runs frogs position´s become closer in each memeplex. Indeed, this problem leads to a premature convergence. For getting better effiency we propose a novel search strategy by infecting not only the worst indvidual but also the best indvidual idea´s. To further improve the speed of convergence of the algorithm, we have introduced two acceleration factors in the search strategy formulation. The proposed algorithm has been evaluated on five mathematical benchmark functions. Compared with a the orginal SFLA and the particle swarm optimization algorithm, the experimental results in terms of optimization performance and the speed of convergence shows that the proposed algorithm can be an effective tool for solving combinatorial optimization problems.