Recent research shows that, in addition to the universal approximation function, neural networks are capable of accurately approximating the mapping, both linear and non-linear, of one space of continuous functions into another. We report on our experience of coupling a well known CFD code, OpenFOAM, to neural network modelling using TensorFlow on a HPC system in order to model tank transfer functions. The CFD stage is well suited to using multiple CPUs via MPI libraries while the machine learning stage is better suited to using GPUs via several Python packages and libraries. We report that these distinct steps can be joined into an efficient pipeline by using the file system effectively. Our results show that neural network modelling works well in the region where non-linear CFD theories are needed to model the dynamics of water waves.
|Title of host publication||IEEE 17th International Conference on eScience (eScience 2021): Proceedings|
|Publisher||Institute of Electrical and Electronics Engineers Inc.|
|Number of pages||6|
|Publication status||Published - 26 Oct 2021|
|Event||17th IEEE International Conference on eScience, eScience 2021 - Virtual, Online, Austria|
Duration: 20 Sep 2021 → 23 Sep 2021
|Name||IEEE International Conference on eScience (eScience 2021): Proceedings|
|Conference||17th IEEE International Conference on eScience, eScience 2021|
|Period||20/09/2021 → 23/09/2021|
Bibliographical noteFunding Information:
The Bryden Centre project is supported by the European Union’s INTERREG VA Programme, managed by the Special EU Programmes Body (SEUPB). The Kelvin-2 HPC centre in Northern Ireland, located at Queen’s University Belfast, and where most of the computations were performed, is funded by the UK EPSRC, grant number EP/T022175/1.
OpenFOAM is an open source HPC software package developed by OpenCFD. It is widely used in science and engineering by both commercial and academic organisations. OpenFOAM shares with other HPC suites the properties that it is written in C++ and that it uses MPI to run in parallel on multiple CPUs. Most recent advances in improving the computational efficiency have been reported by improving the algebraic multigrid algorithm  or the use of higher order temporal schemes . A project recently funded by the European Commission, under its Horizon 2020 Programme aims to increase performance for extreme scale simulations https://exafoam.eu/.
© 2021 IEEE.
ASJC Scopus subject areas
- Computer Science (miscellaneous)
- Artificial Intelligence
- Computer Networks and Communications
- Hardware and Architecture
- Information Systems and Management
- Safety, Risk, Reliability and Quality