Exploring the link between emulation and optimization for the robust design of complex engineered products
Lead Research Organisation:
University of Sheffield
Department Name: Automatic Control and Systems Eng
Abstract
A major challenge for the use of high-fidelity simulations in the design
workflows of complex engineered products is the resource limitations on the number
of times that the model can be run. This limits both the ability to calibrate the
simulation to real-world data and to perform optimization using the calibrated model.
Both calibration and optimization rely on the use of emulators - lower-fidelity
representations of the high-fidelity simulations - but these uses are decoupled in the
overall design workflow. This PhD will explore the relationship between emulation
and optimization processes and aim to develop novel approaches that combine the
two processes into an efficient algorithm framework for design optimization on a
budget. The methods will be shaped and tested within Theme 8 of the Programme
for Simulation Innovation (PSi) funded by EPSRC and Jaguar Land Rover.
workflows of complex engineered products is the resource limitations on the number
of times that the model can be run. This limits both the ability to calibrate the
simulation to real-world data and to perform optimization using the calibrated model.
Both calibration and optimization rely on the use of emulators - lower-fidelity
representations of the high-fidelity simulations - but these uses are decoupled in the
overall design workflow. This PhD will explore the relationship between emulation
and optimization processes and aim to develop novel approaches that combine the
two processes into an efficient algorithm framework for design optimization on a
budget. The methods will be shaped and tested within Theme 8 of the Programme
for Simulation Innovation (PSi) funded by EPSRC and Jaguar Land Rover.
Organisations
People |
ORCID iD |
Robin Purshouse (Primary Supervisor) | |
Oliver Jones (Student) |
Publications
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/N509735/1 | 30/09/2016 | 29/09/2021 | |||
1803697 | Studentship | EP/N509735/1 | 30/09/2016 | 29/09/2019 | Oliver Jones |
Description | This work has laid out a combined mathematical formulation for the joint problem of performing model calibration and optimisation. It has put forward both altered test functioned from within the optimisation literature as well as deriving a new module to work with the well know WFG test problem suite that can be used to assess the performance of the joint problem. A new methodology looking at solving the combined problem has been developed. This method was further developed to improve its efficiency. Both of the developed methods were tested against the current practice to asses their viability. Positive results were observed. |
Exploitation Route | This work could potentially help reduce the costs of carrying out the virtual engineering workflow for cases where expensive complex systems/models are being looked at. It could be used as a baseline/ benchmark for future development and study into the combined problem of optimisation and model calibration. The new methods could be developed further possibly through through the incorporation of efficient robust methodologies. |
Sectors | Manufacturing, including Industrial Biotechology |
Description | SYNERGY Summer School on EFFICIENT MULTI-OBJECTIVE OPTIMISATION |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | Summer school with the aim of expand knowledge regarding Multi objective optimisation and its state of the art practices. Allow for networking with colleges from other countries and provide an opportunity to propose and tackle that they faced within their work. |
Year(s) Of Engagement Activity | 2018 |
URL | http://summerschool.synergy-twinning.eu/ |