Elephant Herding Optimization for Benchmark Problems: Algorithm Design, Implementation, and Performance Analysis
Year: 2025
Type: article
Abstract: This paper presents an in-depth study on the development and evaluation of an Elephant Herding Optimization (EHO) algorithm tailored to solve a set of standard benchmark functions, specifically f1 through f10. Drawing inspiration from the social behavior of elephant herds, the EHO algorithm employs strategies that mimic the leadership of matriarchs, the exploration conducted by male elephants, and the dynamic interplay between diversification and intensification during search. The study begins with a review of contemporary metaheuristic algorithms used in optimization tasks—such as Genetic Algorithm (GA), Differential Evolution (DE), Particle Swarm Optimization (PSO), Simulated Annealing (SA), Firefly Algorithm (FFA), Cuckoo Search (CS), and Tree Physiology Optimization (TPO)—and highlights existing performance gaps when solving complex benchmark functions. We then outline the design of the basic EHO method and introduce improved variants that incorporate novel individual updating strategies. These improvements include replacing suboptimal individuals with higher fitness particles, increasing the number of male elephants to enhance exploratory capabilities, and incorporating information from previous iterations to accelerate convergence. The proposed algorithm is benchmarked on ten classic test functions to comprehensively evaluate its convergence behavior, computational efficiency, and ability to search for global optima. Comparative analysis shows that the improved variants of EHO not only achieve faster convergence but also demonstrate enhanced robustness in terms of statistical consistency across multiple runs. The paper concludes with a discussion on the implications of these findings for both large-scale engineering applications and further research on nature-inspired optimization techniques.
DOI: 10.14419/vc2c7252
Institution: Akre University for Applied Sciences
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4415544683