Rüppell’s fox optimizer: A novel meta-heuristic approach for solving global optimization problems

A new swarm-based, nature-inspired meta-heuristic method, called Rüppell’s Fox Optimizer (RFO), is designed and examined in this study to address global optimization problems. RFO takes inspiration from the natural and intelligent communal foraging practices of Rüppell’s foxes, both during the day and at night. The optimizer mathematically simulates a variety of chief foraging activities of Rüppell’s foxes using their acute eyesight, hearing, and scent to hunt, to accommodate both exploration and exploitation aspects throughout the optimization action. The RFO algorithm simulates the rotating eyes’ feature of Rüppell’s foxes to a field of view of about 260°, as well as the feature of rotating their ears to a field of hearing of about 150°. Further, a cognitive component, ρ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\rho $$\end{document}, is designed to oversee the evolution from global search to local search as well as the equilibrium between exploration and exploitation within the search domain. Based on Rüppell’s foxes’ foraging qualities, RFO reveals a variety of foraging habits. The dynamic patterns and activities of Rüppell’s foxes were mathematically mimicked in this study to create an efficient global optimizer, which when applied to optimization problems, ultimately yields feasible solutions. The performance level of RFO was thoroughly examined through a comparison with 12 other algorithms using three challenging test sets: Congress on Evolutionary Computation 2017 (CEC 2017) group with 29 test functions of various dimensions, the Congress on Evolutionary Computation 2019 (CEC 2019) with 10 test functions, and CEC 2020 set with 10 test functions. Using mean rank of Friedman’s test, RFO outperformed other comparative algorithms by 31.25%, 34.04%, 32.36%, and 55.13% on the 10-, 30-, 50-, and 100-dimensional CEC 2017 test set, respectively, 57.11% on the CEC 2019 test suite with 10 dimensions, 48.59%,and 51.68% on the 10- and 20-dimensional CEC 2020 set, respectively. The applicability of RFO is thoroughly examined on 6 classic engineering design problems and a wind farm layout problem under several constraints . The outcomes illustrate the ascendancy and encouraging potential of RFO when tackling a broad range of difficult real-world problems. All things considered, the proposed algorithm performs exceptionally well with respect to exploitation, exploration, striking a balance between the former two aspects, and avoiding local optima. RFO is quite competitive, especially for optimization problems with many constraints and variables, as well as unimodal and multimodal features. As per Friedman’s test, which is followed up by Holm’s test, RFO considerably beats all other competing algorithms for the bulk of the examined test functions.