Multi-population following behavior-driven fruit fly optimization: A Markov chain convergence proof and comprehensive analysis

Fruit fly optimization algorithm (FOA) is a well-known optimization algorithm with a well-designed structure and superiority of fewer parameters, more effortless adjustment, and competitive and fast computation time. Up till now, FOA has been effectively applied to numerous fields, such as financial forecast and medicine, and it achieves favorable results. However, some aspects need to be enhanced when dealing with some function optimization cases. This method is inclined to falling into local optima with a slow convergence. In this paper, the following behavior-driven multi-population FOA is proposed to relieve these drawbacks, which combines the following mechanism of artificial fish swarm algorithm with the unique searching ability of different types of fruit flies. The chaotic global disturbance is introduced to improve the global exploration ability of the original FOA and reduce the probability of advanced FOA falling into the local extreme. The proposed FOA, it is compared with the advanced FOA with several well-established algorithms and the latest improved optimizers in different dimensions horizontally and vertically to substantiate the effectiveness of the. It is also applied advanced FOA to two practical engineering optimization projects. Systematic analysis and experimental data indicate that the advanced FOA variant outperforms the original FOA and the latest improved algorithms. An online repository will support this research at http://aliasgharheidari.com for any communication and guidance for future works. (C) 2020 Elsevier B.V. All rights reserved.