You Guoping, Hu Yudan, Yang Zhen, Li Yuhang
School of Communications and Electronics, Jiangxi Science and Technology Normal University, Nanchang, China.
School of Foreign Languages, Zhijiang College of Zhejiang University of Technology, Shaoxing, China.
Sci Rep. 2025 Mar 23;15(1):10036. doi: 10.1038/s41598-025-93410-w.
The snow ablation optimizer (SAO) represents a novel metaheuristic algorithm tailored for addressing real-world optimization challenges. However, SAO exhibits certain drawbacks, including a tendency to get trapped in local optima, a sluggish convergence rate, and suboptimal performance on intricate multimodal function problems. Acknowledging these limitations, the enhanced snow ablation optimizer (ESAO) is introduced. In this paper, we elucidate the pivotal strategies for implementing ESAO, encompassing chaotic mapping and random opposition learning initialization, dynamic tangential flight strategy, adaptive inertia weight, and elite guidance boundary control strategy. To underscore the prowess of ESAO, we conducted extensive testing on 29 functions from the CEC2017 benchmark, 19 real-world engineering challenges derived from the CEC2020 benchmark functions, and UAV flight trajectory optimization. Furthermore, ESAO is compared with three categories of widely recognized algorithms: (1) classical algorithms such as PSO, HHO, and GWO; (2) recent algorithms like GOOSE, HEOA, Puma, and the original SAO; and (3) algorithmic variants including IGWO, IDBO, HPHHO, and E-WOA. The experimental outcomes reveal that ESAO surpasses the other 11 competitors in most cases, demonstrating remarkable convergence speed, stability, and accuracy. The superiority of ESAO is further confirmed by the Friedman mean ranking test and Wilcoxon rank sum test, underscoring its potential as a formidable metaheuristic algorithm.
雪消融优化器（SAO）是一种专门为应对现实世界优化挑战而设计的新型元启发式算法。然而，SAO存在一些缺点，包括容易陷入局部最优、收敛速度缓慢以及在复杂多峰函数问题上性能欠佳。认识到这些局限性后，提出了增强型雪消融优化器（ESAO）。在本文中，我们阐述了实现ESAO的关键策略，包括混沌映射和随机反向学习初始化、动态切线飞行策略、自适应惯性权重以及精英引导边界控制策略。为了突出ESAO的优势，我们对来自CEC2017基准的29个函数、从CEC2020基准函数衍生的19个实际工程挑战以及无人机飞行轨迹优化进行了广泛测试。此外，还将ESAO与三类广泛认可的算法进行了比较：（1）经典算法，如PSO、HHO和GWO；（2）近期算法，如GOOSE、HEOA、Puma以及原始的SAO；（3）算法变体，包括IGWO、IDBO、HPHHO和E-WOA。实验结果表明，在大多数情况下ESAO优于其他11个竞争对手，展现出显著的收敛速度、稳定性和准确性。Friedman均值排序检验和Wilcoxon秩和检验进一步证实了ESAO的优越性，凸显了其作为一种强大元启发式算法的潜力。
