An Improved Snow Ablation Optimizer Based on Multi-Strategy Fusion for Global Optimization Problems

The Snow Ablation Optimizer (SAO) is a novel meta-heuristic search algorithm proposed in 2023, which solves complex optimization problems by simulating snow's melting and sublimation processes. However, SAO experiences slow convergence rates and is prone to getting stuck in local optima. which limits its performance. To address these issues, an Improved Snow Ablation Optimizer (ISAO) is proposed in this paper. First, Latin Hypercube Sampling (LHS) is used in the population initialization stage, which significantly improves the diversity of the initial population and the coverage of the search space. Second, the exploration phase incorporates the Whale Optimization Algorithm (WOA) with combined encircling attack and spiral update strategies, enhancing global search capability and effectively avoiding local optima. Finally, Variable Neighborhood Search (VNS) is introduced in the exploitation phase, and three neighborhood structures, namely, micro-perturbation, strong-perturbation, and jump-perturbation, are designed to achieve a more flexible and refined local search. To validate the performance of ISAO, this paper conducts experimental evaluations on the CEC2017 benchmark test functions, including comparative analyses with the original algorithm and its three improved modules, as well as performance comparisons with five classical optimization algorithms. In addition, the Wilcoxon rank sum test was employed to verify the significance of ISAO. Experimental results show that ISAO significantly outperforms the original and comparison algorithms regarding convergence speed, solution accuracy, and robustness, demonstrating excellent optimization performance.