Two-Phases Learning Shuffled Frog Leaping Algorithm

In order to overcome the drawbacks of standard shuffled frog leaping algorithm that converges slowly at the last stage and easily falls into local minima, this paper proposed two-phases learning shuffled frog leaping algorithm. The modified algorithm added the elite Gaussian learning strategy in the global information exchange phase, updated frog leaping rule and added the learning capability that the worst frog of current swarm learned from the best frog of other swarm. The learning capability of two-stage on the one hand increased the search range, on the other hand enhanced the diversity of population. Experiments were conducted on 13 classical benchmark functions, the simulation results demonstrated that the proposed approach improved the convergence rate and solution accuracy, when compared with common swarm intelligence algorithm and the latest improved shuffled frog leaping algorithm.