Bespoke and continually developing optimisation methodologies are central to our tool sets. Over 10 years of development, this has become one of the key performance differentiators between ourselves and ‘traditional’ marine engineering offices.
Contemporary vessel design is a far more complex procedure now than in recent decades and made more complex when considering the ship as a holistic system – a must if ultimate efficiency is to be achieved.
At BAR we have long since changed our approach to one using machine learning and training neural networks to compute the best hull formation when considering all of the objectives.
We are now designing the systems to design the hull…
- One tool doesn’t suit all- flexible tool building blocks which allows quick creation of specific optimisation tool
- Large-scale Batch Creation of Parametric Geometries
- Automated and Highly Accurate Simulation Techniques – Primarily RANS CFD based.
- Optimisation Techniques including Progressive Neural Network/AI Techniques which aids multi-objective no stone unturned approach
- Proven technique and approach for the optimisation of hulls
- Typically around 15 parameter networks, 10~12 of which are geometrical
- Automated geometric creation and RANs CFD
- AI derived hull shapes
- Multi-objective, weighted performance profile approach – No cliffs