Dynamic Chaotic Opposition-Based Learning-Driven Hybrid Aquila Optimizer and Artificial Rabbits Optimization Algorithm: Framework and Applications

Abstract

Aquila Optimizer (AO) and Artificial Rabbits Optimization (ARO) are two recently developed meta-heuristic optimization algorithms. Although AO has powerful exploration capability, it still suffers from poor solution accuracy and premature convergence when addressing some complex cases due to the insufficient exploitation phase. In contrast, ARO possesses very competitive exploitation potential, but its exploration ability needs to be more satisfactory. To ameliorate the above-mentioned limitations in a single algorithm and achieve better overall optimization performance, this paper proposes a novel chaotic opposition-based learning-driven hybrid AO and ARO algorithm called CHAOARO. Firstly, the global exploration phase of AO is combined with the local exploitation phase of ARO to maintain the respective valuable search capabilities. Then, an adaptive switching mechanism (ASM) is designed to better balance the exploration and exploitation procedures. Finally, we introduce the chaotic opposition-based learning (COBL) strategy to avoid the algorithm fall into the local optima. To comprehensively verify the effectiveness and superiority of the proposed work, CHAOARO is compared with the original AO, ARO, and several state-of-the-art algorithms on 23 classical benchmark functions and the IEEE CEC2019 test suite. Systematic comparisons demonstrate that CHAOARO can significantly outperform other competitor methods in terms of solution accuracy, convergence speed, and robustness. Furthermore, the promising prospect of CHAOARO in real-world applications is highlighted by resolving five industrial engineering design problems and photovoltaic (PV) model parameter identification problem.