1. Applying Neural Networks to Predict Offshore Platform Dynamics
- Author
-
Nikolas Martzikos, Carlo Ruzzo, Giovanni Malara, Vincenzo Fiamma, and Felice Arena
- Subjects
artificial neural networks ,offshore platforms ,aquaculture platforms ,mooring loads ,renewable energy ,Naval architecture. Shipbuilding. Marine engineering ,VM1-989 ,Oceanography ,GC1-1581 - Abstract
Integrating renewable energy sources with aquaculture systems on floating multi-use platforms presents an innovative approach to developing sustainable and resilient offshore infrastructure, utilizing the ocean’s considerable potential. From March 2021 to January 2022, a 1:15-scale prototype was tested in Reggio Calabria, Italy, which gave crucial insights into how these structures behave under different wave conditions. This study investigates the application of Artificial Neural Networks (ANNs) to predict changes in mooring loads, particularly at key points of the structure. By analyzing metocean data, several ANN models and optimization techniques were evaluated to identify the most accurate predictive model. With a Normalized Root Mean Square Error (NRMSE) of 1.7–4.7%, the results show how ANNs can effectively predict offshore platform dynamics. This research highlights the potential of machine learning in developing and managing sustainable ocean systems, setting the stage for future advancements in data-driven marine resource management.
- Published
- 2024
- Full Text
- View/download PDF