Refraction-Learning Based Whale Optimization Algorithm with Opposition-Learning and Adaptive Parameter Optimization

In this paper, the development of an enhanced variant of the Whale Optimization Algorithm (WOA) is proposed which addresses some of the limitations, namely slow convergence speed, a tendency to get stuck in local optima, and loss of population diversity. The proposed variant integrates novel strategies namely Refraction Learning (RL) and Opposition-Based Learning (OBL) along with an adaptive parameter optimization mechanism. Owing to the unique amalgamation of the above-mentioned strategies, the performance of the conventional WOA is augmented by intensifying the search process and smoothing it overall. This leads to superior performance in the case of multi-modal landscapes, escaping local optima, and a more eclectic investigation of the search space. Through comprehensive benchmark testing across 23 functions, demonstration occurs that the algorithm outperforms the original WOA and other modern optimization techniques. The proposed WOA variant shows significant improvements in terms of convergence rate, global search ability, and solution accuracy. The empirical results highlight the strengths of the approach and indicate its potential as an effective optimization tool for solving complex real-world problems.