水下无人系统学报2026，Vol.34Issue(4)：616-626,11.DOI:10.11993/j.issn.2096-3920.2026-0014
基于物理驱动与不确定性量化的海上漂流浮标轨迹溯源方法
Physics-Driven Trajectory Backtracking Method for Drifting Buoys Based on Uncertainty Quantification
摘要
Abstract
High-precision trajectory backtracking technology for drifting buoys is urgently needed for maritime search and rescue and pollution source tracing,yet traditional Lagrangian models exhibit significant errors in complex marine environments.A physics-driven trajectory backtracking model was proposed for marine drifting buoys.A dynamic diffusion coefficient based on autocorrelation analysis of wind flow field time series was innovatively introduced to optimize the subgrid velocity compensation mechanism in random walk models.The model integrates a drift kinetics framework incorporating wind driving force,ocean current drag force,and Coriolis force,and combines Monte Carlo simulation with kernel density estimation to quantify the spatio-temporal uncertainties in trajectory backtracking.Based on buoy data from the North Atlantic Ocean Internet of Things,four typical marine environments were selected,namely,tropical open ocean,current convergence zones,temperate westerlies,and nearshore complex terrains,to complete 72-hour trajectory backtracking validation.According to the result,the proposed model achieves 72-hour trajectory backtracking errors of only 3.9～5.8 km.Compared with the traditional Lagrangian models,accuracy improvements of 74%,55%,59%,and 22%are achieved in the four sea areas,respectively.The trajectory backtracking accuracy in complex marine environments is significantly improved,providing reliable technical support for maritime emergency rescue and pollution source localization.关键词
漂流浮标/轨迹溯源/动态扩散系数/核密度估计Key words
drifting buoy/trajectory backtracking/dynamic diffusion coefficient/kernel density estimation分类
军事科技引用本文复制引用
李慧,王水,熊鑫泉,越鹏,莫李弘,万旺华..基于物理驱动与不确定性量化的海上漂流浮标轨迹溯源方法[J].水下无人系统学报,2026,34(4):616-626,11.基金项目
国家自然科学基金(62403158) （62403158）
中央高校科研业务费(3072024LJ0405) （3072024LJ0405）
黑龙江省春雁人才团队支持计划(CYQN24070). （CYQN24070）
