Dynamic load torque inversion and environmental effect analysis of polar ship anchor windlasses based on improved Grey Wolf Optimizer-driven ensemble learning

2026-08-10

PRE

Haiming Wen

Gang Wu

Meng Cui

Chenyang Liu

Pengcheng Li

Yan Chen

Wei Sun

Dayong Zhang *

DOI：10.1016/j.oceaneng.2026.127498

Original

Original request for help

Save

Abstract

En 中文

• A data-driven framework is developed for dynamic load inversion of polar anchor windlasses. • An I-GWO-optimized Stacking model is proposed for time-domain load inversion. • Coupled simulations and polar-simulated tests validate the model accuracy and robustness. • Ice accretion is quantitatively identified as the dominant environmental factor. • SHAP analysis reveals pressure and temperature as the most influential features.

Keywords:

Polar ship

Anchor windlass

Dynamic load inversion

Ensemble learning

Interpretability analysis

Dynamic load torque inversion and environmental effect analysis of polar ship anchor windlasses based on improved Grey Wolf Optimizer-driven ensemble learning

Abstract

Journal

Organization

Cited Papers

Citing Papers