DEMOPS - Machine learning based speed-power performance modelling to reduce fuel cost and emissions from shipping
A ship's fuel consumption can be significantly increased when sailing in harsh sea conditions. Any measures to increase ship energy efficiency must rely on accurate description of the ship's energy performance, ie, power-speed relationship, at sea.
Current theoretical physical models always contain large uncertainties to describe a ship's energy performance especially in the mechanical system models. Some black-box performance models have been constructed by machine learning methods based on ship performance data. But the black box models can only be useful for a specific ship with data inputted for the model construction.
In this project, we will develop functional data analysis (FDA) algorithms to choose / simulate proper ship data sample.The data will be used for sophisticated machine learning algorithms to combine with theoretical models to better understand and construct ship energy performance models. Some inversed machine learning algorithms will be developed to properly describe a ship's encountered wave environments. The wave conditions will be used to best model the ship's actual performance at sea. Finally, these models will be demonstrated to show how they can be used to develop ship energy efficient measures. Potential fuel savings and air emission reduction will be identified through the demonstrations.
- Molflow AB (Private, Sweden)
- GoTa Ship Management AB (Private, Sweden)
- Lund University (Academic, Sweden)
- SSPA Sweden AB (Private, Sweden)
This page is only available in english
- Swedish Transport Administration (Public, Sweden)
- Lighthouse (Centre, Sweden)