首页|期刊导航|大气和海洋科学快报（英文版）|Tropical cyclone intensity prediction based on Kolmogorov-Arnold networks with predictor pruning optimization

Tropical cyclone intensity prediction based on Kolmogorov-Arnold networks with predictor pruning optimization

Keyun Li Wei Zhong Yao Yao Fangzhao Li Yuan Sun Hongrang He

大气和海洋科学快报（英文版）2026，Vol.19Issue(4)：47-52,6.

大气和海洋科学快报（英文版）2026，Vol.19Issue(4)：47-52,6.DOI:10.1016/j.aosl.2025.100694

Tropical cyclone intensity prediction based on Kolmogorov-Arnold networks with predictor pruning optimization

Keyun Li ¹Wei Zhong ²Yao Yao ³Fangzhao Li ²Yuan Sun ²Hongrang He²

作者信息

1. College of Meteorology and Oceanography,National University of Defense Technology,Changsha,China
2. College of Advanced Interdisciplinary Studies,National University of Defense Technology,Changsha,China
3. College of Advanced Interdisciplinary Studies,National University of Defense Technology,Changsha,China||School of Atmospheric Sciences,Nanjing University,Nanjing,China
折叠

摘要

Abstract

准确的热带气旋(TC)强度预报对减灾和公共安全至关重要.然而,目前TC强度预报存在因子筛选复杂,预报精度不足等问题.为此,本文构建了具备因子剪枝优选能力的Kolmogorov-Arnold网络(KANs)全球TC强度智能预报模型(TCI-KAN).该模型设计了数据驱动的预报因子筛选方法,通过权重排序分析实现了低影响预报因子的迭代剪枝.结果表明,TCI-KAN在6小时TC强度预报中表现优异,独立测试集平均绝对误差(MAE)为2.85 kt,较美国国家飓风中心官方预报,最优单深度学习模型及最优混合模型分别降低31％,13％和6％.进一步分析表明,TCI-KAN适用于不同海域和TC强度类别.

关键词

热带气旋/Kolmogorov-Arnold网络/预报因子剪枝优选/强度预报

Key words

Tropical cyclone/Kolmogorov-Arnold networks/Predictor pruning optimization/Intensity prediction

引用本文复制引用

Keyun Li,Wei Zhong,Yao Yao,Fangzhao Li,Yuan Sun,Hongrang He..Tropical cyclone intensity prediction based on Kolmogorov-Arnold networks with predictor pruning optimization[J].大气和海洋科学快报（英文版）,2026,19(4):47-52,6.

基金项目

This research was supported by the National Natural Science Foun-dation of China[grant numbers 42075011 and 42192552]. （）

大气和海洋科学快报（英文版）

ISSN：1674-2834

访问量0

下载量0

段落导航