Firefly Algorithm Based on Euclidean Metric and Dimensional Mutation

Abstract

Firefly algorithm is a meta-heuristic stochastic search algorithm with strong robustness and easy implementation. However, it also has some shortcomings, such as the “oscillation” phenomenon caused by too many attractions, which makes the convergence speed too slow or premature. In the original FA, the full attraction model makes the algorithm consume a lot of evaluation times, and the time complexity is high. Therefore, in this paper, a novel firefly algorithm (EMDmFA) based on Euclidean metric (EM) and dimensional mutation (DM) is proposed. The EM strategy makes the firefly learn from its nearest neighbors. When the firefly is better than its neighbors, it learns from the best individuals in the population. It improves the FA attraction model and dramatically reduces the computational time complexity. At the same time, DM strategy improves the ability of the algorithm to jump out of the local optimum. The experimental results show that the proposed EMDmFA significantly improves the accuracy of the solution and better than most state-of-the-art FA variants.