Optimisation of Ship Form Based on Seakeeping Behaviour Using Machine Learning


In this talk, presented by Pablo Romero-Tello and team, the central topic was the analysis of seakeeping behaviour, the study of the movements and forces produced by waves in marine systems. This analysis is of paramount importance in naval design as key parameters like the operability of the ship, passenger comfort, propulsion performance, manoeuvrability, and the operability of equipment and systems heavily depend on it.

Historically, seakeeping behaviour has been analysed using tests on hydrodynamic experience channels or with numerical models. However, recent years have witnessed the advent of Artificial Intelligence in this domain. Several research works have emerged, proposing the use of Machine Learning (ML) techniques for studying the seakeeping behaviour of ships. In this work, the use of a pre-trained Artificial Neural Network (ANN) was highlighted for predicting seakeeping behaviour. Leveraging the speed of ML-based predictions, a substantial number of ships can be analysed in reduced computation time compared to traditional techniques.

The main objective presented was the search for ship geometry best adapted to specific sea conditions and operational profiles by optimising specific metrics related to their operability. The proposed solution was ship hull form optimisation using innovative techniques like Genetic Algorithms (GA) and Particle Swarm Optimization (PSO), which are linked to the surrogate ANN solver developed earlier in the research.

The talk concluded by presenting the most pivotal findings and implications of the work. Attendees were part of the session “IS02 (II) - Artificial Intelligence Applied in Marine Engineering” at the prestigious 10th International Conference on Computational Methods in Marine Engineering (MARINE 2023).

The research for this project is funded by BBVA Foundation and Agencia Estatal de Investigación.