Shuffled frog-leaping algorithm using elite opposition-based learning

Shuffled frog-leaping algorithm (SFLA) has been shown that it can yield good performance for solving various optimisation problems. However, it tends to suffer from premature convergent when solving complex problems. This paper presents an effective approach, called SFLA using elite opposition-based learning (EOLSFLA), which employs elite population to overcome the problems through adopting opposition-based learning to generate opposite solution, increasing the capability of neighbourhood search and enhancing the local exploitation ability. At the same time, to improve the learning ability of the worst population in a memeplex, a new frog-leaping rule is proposed, in which the worst population can learn from other population in the memeplex to reinforce the global space-exploration ability. Experiments are conducted on a set of well-known benchmark functions to verify the performance of EOLSFLA, comparing with other standard swarm intelligence algorithms, opposition-based learning algorithms and improved SFLA, the results demonstrate promising performance of the new method EOLSFLA on convergence velocity and precision.