A chaotic exploitation enhancement for African Buffalo Optimisation algorithm: Empirical evaluation on standard benchmark functions
Year: 2026
Type: article
Abstract: The African Buffalo Optimisation (ABO) algorithm is a nature-inspired metaheuristic that has demonstrated competitive performance on combinatorial optimisation problems. However, the classical ABO algorithm suffers from a fundamental limitation in its exploitation process, which relies on manually assigned control parameters determined through trial-and-error approaches. This arbitrary parameter selection often results in suboptimal exploitation, causing the algorithm to converge prematurely to local optima rather than the global optimum. This paper proposes a chaotic exploitation enhancement for ABO, denoted as expltABO, which replaces the static exploitation parameters with a Tent-map chaotic function that adaptively tunes the exploitation intensity throughout the optimisation process. The proposed algorithm is rigorously evaluated against five benchmark algorithms—PSO, ACO, CS, GA, and classical ABO—on a comprehensive suite of 16 standard optimisation functions (F1–F16) encompassing unimodal, multimodal, and hybrid landscapes. Empirical results demonstrate that expltABO consistently outperforms all benchmark algorithms across the majority of test functions, achieving the lowest error values with superior convergence characteristics. The Tent-map chaotic adaptation mechanism enables dynamic adjustment of the exploitation–exploration balance, preventing premature convergence while maintaining solution precision. Statistical analysis using the Wilcoxon signed-rank test confirms the significance of the improvements (p<0.05). The findings establish chaotic exploitation tuning as an effective enhancement strategy for ABO, with particular applicability to problems requiring precise local search capabilities.
Institution: Hussaini Adamu Federal Polytechnic
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7202186554