Scripts and data for application of the Adaptive Screening method to three applications

This set of scripts and data files can be used to re-generate the Adaptive Screening method and the three applications described in the paper "Designing for dangerous waves – definition and application of a new ‘Adaptive Screening’ method to predict extreme values of non-linear ship responses to waves".<br>Predicting extreme values of strongly non-linear ship responses (such as wave impact loads) is crucial for ensuring safety and performance of maritime structures. However, this is challenging due to the complexity and rarity of the responses. Existing methods are limited, as they are either suitable for weakly non-linear responses only, or are very computationally intensive. The paper above in combination with this dataset introduces a new event-based multi-fidelity method called ‘Adaptive Screening’ to efficiently predict extreme values of strongly non-linear wave-induced responses. It combines elements of screening, multi-fidelity Gaussian Process Regression, and adaptive sampling. Three applications validate the effectiveness of the new method in the paper: one weakly non-linear case where we predict extreme values of second-order waves, one intermediate case where we predict extreme values of vertical bending moments, and one strongly non-linear case where we predict extreme values of green water impact loads. The input necessary to reproduce applications 1 and 2 is also included in the dataset (application 3 relies on proprietary data, which are not part of this repository).<br>The paper demonstrates that Adaptive Screening outperforms conventional brute-force methods, achieving comparable accuracy in predicting extreme values while significantly reducing high-fidelity simulation times (especially for the most non-linear cases). Like many alternative methods, Adaptive Screening relies on a response-dependent low-fidelity indicator variable. We also show that the method performs well with realistic indicators for a range of applications. The test cases indicate that Adaptive Screening is very promising for the strongly non-linear responses it was designed for.

本脚本与数据文件集可用于复现论文《面向危险波浪的设计——新型“自适应筛选（Adaptive Screening）”方法的定义与应用：预测波浪作用下船舶非线性响应的极值》中所述的自适应筛选方法及三项应用案例。
针对强非线性船舶响应（如波浪冲击载荷）的极值预测，是保障海事结构安全与运行性能的核心环节。但由于此类响应机理复杂且发生概率极低，该任务极具挑战性。现有方法存在显著局限：要么仅适用于弱非线性响应场景，要么计算开销极高。上述论文结合本数据集，提出了一种全新的基于事件的多保真度方法——自适应筛选，可高效实现强非线性波浪诱导响应的极值预测。该方法融合了筛选分析、多保真度高斯过程回归（Gaussian Process Regression）与自适应采样技术。论文中通过三项应用验证了该方法的有效性：其一为弱非线性场景，用于预测二阶波浪的极值；其二为中等非线性场景，用于预测垂向弯矩的极值；其三为强非线性场景，用于预测上浪冲击载荷（green water impact loads）。数据集已包含复现前两项应用所需的全部输入数据（第三项应用依赖专有数据，未纳入本代码仓库）。
研究表明，自适应筛选方法的性能优于传统蛮力法（brute-force methods）：在极值预测任务中可达到相当的精度，同时大幅缩短高保真仿真时长（尤其针对强非线性场景）。与诸多同类替代方法类似，自适应筛选依赖于与响应相关的低保真度指示变量。此外，研究验证了该方法在多款实际应用场景的指示变量下均表现优异。测试案例显示，自适应筛选在其针对的强非线性响应场景中极具应用前景。