Estimating semi-stationary sea states through wavelet-based adaptive segmentation of ship motions
Published in Ocean Engineering, 2026
为提升复杂海上作业的安全性，本文聚焦非平稳环境下波浪浮标类比（WBA）的局限性。传统 WBA 通常在固定 30 分钟时间窗内假设海浪平稳，但真实海浪具有明显的时间演化特征，忽略该特性会导致显著估计误差。本文提出“半平稳海况”的概念，并采用基于小波的自适应分段方法，在船舶运动响应谱中识别环境转变点。数值仿真验证了该算法能够优化时频权衡，从而避免病态逆问题中扰动被放大的现象。尽管当前研究使用理想的传递函数（RAO），该框架为时间相关海况估计提供了稳健机制，并指出未来可通过受限等距性质等准则进一步应对 RAO 不确定性。
To improve the safety of complex marine operations, this research tackles the limitations of the wave buoy analogy (WBA) in non-stationary environments. While traditional WBA assumes stationarity over fixed 30-minute windows, real-world waves exhibit a time-involved nature that leads to significant estimation errors when ignored. We introduce a novel framework that defines semi-stationary sea states and utilizes wavelet-based adaptive segmentation to identify environmental transitions in ship motion response spectra. Validated through numerical simulations, the algorithm optimizes the time-frequency trade-off, preventing the amplification of disturbances inherent in ill-posed inverse problems. While the current study utilizes ideal transfer functions (RAOs), it establishes a robust mechanism for time-dependent sea state estimation and highlights future research directions in addressing RAO uncertainty through criteria like the restricted isometry property.
Recommended citation: Taiyu Zhang, Can Ma, Yang An, and Zhengru Ren*. (2026). "Estimating semi-stationary sea states through wavelet-based adaptive segmentation of ship motions." Ocean Engineering. 350, 124030.
Download Paper