Towards probabilistic models for the prediction of a ship performance in dynamic ice
Research output: Contribution to journal › Article › Scientific › peer-review
This paper introduces two probabilistic, data-driven models that predict a ship's speed and the situations where a ship is likely to get stuck in ice based on the joint effect of ice features such as the thickness and concentration of level ice, ice ridges, rafted ice, moreover ice compression is considered.
To develop the models, two full-scale datasets were utilized. First, the dataset about the performance of a selected ship in ice is acquired from the automatic identification system. Second, the dataset containing numerical description of the ice field is obtained from a numerical ice model HELMI, developed in the Finnish Meteorological Institute.
The collected datasets describe a single and unassisted trip of an ice-strengthened bulk carrier between two Finnish ports in the presence of challenging ice conditions, which varied in time and space.
The relations between ship performance and the ice conditions were established using Bayesian networks and selected learning algorithms.
The obtained results show good prediction power of the models. This means, on average 80% for predicting the ship's speed within specified bins, and above 90% for predicting cases where a ship may get stuck in ice.
|Journal||Cold Regions Science and Technology|
|Publication status||Published - 2015|
|MoE publication type||A1 Journal article-refereed|
- Bayesian networks, Machine learning, Ship beset in ice, Ship performance in ice