Minimal dataset for "Physics-Structured Residual Learning for Ship Maneuvering Prediction: Multi-Source Disturbance Decomposition and Compensation"

This upload contains the minimal dataset supporting the findings of the study “Physics-Structured Residual Learning for Ship Maneuvering Prediction: Multi-Source Disturbance Decomposition and Compensation”. The package includes representative figure files, numerical data for the main tables and trajectory plots, and the core evaluation/plotting scripts used to generate the key figures and summary statistics reported in the manuscript. It is intended as a compact reproducibility package for figure and result-level inspection, rather than a full archive of the complete training pipeline. Additional raw data are available from the corresponding author upon reasonable request.本上传包包含支撑研究《面向船舶操纵预测的物理结构化残差学习：多源扰动分解与补偿》结论所需的最小规模数据集。 该包内含代表性图像文件、主表格与轨迹图对应的数值数据，以及用于生成论文中关键图表与汇总统计量的核心评估与绘图脚本。本包旨在作为轻量化可复现工具包，用于对图表及结果层面进行检视，而非完整训练流程的全量归档。 合理申请后，可向通讯作者获取额外原始数据。