Prediction error statistics in deterministic linear ship motion forecasting
Deterministic ship motions predictions methodologies represent a promising emerging approach, which could be embedded in decision support systems for certain types of operation. The typically envisioned prediction chain starts from the remote sensing of the wave elevation through wave radar technolo...
| Published in: | Volume 7A: Ocean Engineering |
|---|---|
| Main Authors: | , , |
| Other Authors: | |
| Format: | Conference Object |
| Language: | English |
| Published: |
American Society of Mechanical Engineers (ASME)
2018
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/11368/2930954 https://doi.org/10.1115/OMAE2018-77456 http://proceedings.asmedigitalcollection.asme.org/proceeding.aspx?articleid=2704528 |
| Summary: | Deterministic ship motions predictions methodologies represent a promising emerging approach, which could be embedded in decision support systems for certain types of operation. The typically envisioned prediction chain starts from the remote sensing of the wave elevation through wave radar technology. An estimated wave field is then fitted to the data, it is propagated in space and time, and it is finally fed to a ship motion prediction model. Prediction time horizons, typically, are practically limited to the order of minutes. Deterministic predictions are, however, inevitably associated with prediction uncertainty which is seldom quantified. This paper, therefore, presents a semi-analytical methodology for the estimation of ship motion prediction error statistics in ensemble domain as function of the forecasting time, assuming linear Gaussian irregular waves and stationary linear ship motions. This information can be used, for instance, to supplement deterministic forecasting with corresponding confidence intervals. The paper describes the theoretical background of the developed methodology and reports some numerical application examples. |
|---|