Users of Computer Aided Engineering applications always ask for higher computational speed and accuracy.
Adopting Digital Twins broadly in the future, we expect this need to significantly increase.
Recently, hybrid technologies - combining machine learning and classical simulation technologies - have been proposed to bring computational speed and accuracy of simulation tools to a new level.
They thus have the potential to address user needs significantly better.This PhD project seeks to explore a cutting-edge hybrid approach that combines machine learning with classical simulation methods to advance computational speed and accuracy.
Specifically, the project will investigate the integration of Neural Operators-efficient learning-based partial differential equation (PDE) solvers defined on simplified domains (e.g., unit squares)-with domain decomposition strategies.
This hybrid methodology aims to establish a new standard in simulation performance.
We invite applications from highly motivated individuals to join this project and contribute to this exciting area of research.
Applicants should have (or expect to obtain by the start date) at least a high 2.1 degree (preferably a first or its equivalent) in Engineering, Machine learning, Applied Mathematics or related subject.
This studentship is open to both home and overseas applicants.
The successful candidate will work collaboratively with a multidisciplinary team based in Cambridge Universiy and Siemens Digital Industry Software, gaining expertise in advanced computational methods and state-of-the-art machine learning techniques.
EPSRC ICASE studentships are fully-funded (fees and maintenance) for students eligible for Home fees.
EU and international students may be considered for a small number of awards at the Home fees rate.
Full eligibility criteria can be found via the following link; The University actively supports equality, diversity and inclusion and encourages applications from all sections of society.