Ship as a Buoy for Real-time Deterministic Forecast of Nonlinear Waves and Ship Motion via Data Assimilation

Abstract

The objective of this project is to develop the capability of real-time (online) forecast of nonlinear phase-resolved waves and ship motion via simulations coupled by an ensemble-based data assimilation framework. The data coming into play includes radar and/or discrete buoy measurements of surface waves, and, most importantly, the ship itself as a buoy (e.g., measured ship motion can be assimilated). The framework has at least two major novelties compared to previous works in this field: (1) The ensemble-based data assimilation, based on the PI#s recent two published work, is essential to real-time application. This is in contrast to the previous methods using variational data assimilation which is very difficult to implement in real time. (2) Our #ship as a buoy# method is drastically different from this concept used in previous contexts. Here we aim at directly coupling the measured ship motion to a phase-resolved wave field, in contrast to only inferring the wave spectrum as proposed in previous works. Our overall framework, implemented on CPU-GPU hardware and with online data supply, can simultaneously provide the next-minute wave field around the ship, the ship motion (aiming for 6 DoF and assuming ship path is known), as well as parameter corrections in the ship motion model. We will also develop another novel approach to use trained neural work to reduce the required number of samples in the ensemble forecast, i.e., providing additional acceleration to the overall algorithm. The developed method is especially useful in extreme wave conditions where predicting the precise ship motions induced by each individual wave is beneficial.

Document Details

Document Type
DoD Grant Award
Publication Date
Nov 08, 2024
Source ID
N000142412535

Entities

People

  • Yulin Pan

Organizations

  • Board of Regents of the University of Michigan
  • Office of Naval Research
  • United States Navy

Tags

Fields of Study

  • Environmental science

Readers

  • Atmospheric Science/Meteorology
  • Marine Hydrodynamics
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms