Towards Improving the Design Space Exploration Process Using Generative Design With MBSE

Louis R Timperley, Lucy Berthoud, Chris M Snider, Theo Tryfonas, Antonio Prezzavento, Kyle Palmer

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

2 Citations (Scopus)
39 Downloads (Pure)

Abstract

Model Based Systems Engineering (MBSE) is an interesting alternative to traditional systems engineering methods. Instead of using electronic documents to record system information, MBSE uses a unified and coherent system model. Trade-offs are a major element of a space systems engineer’s role in early system design. This can be a particularly challenging process in the domain of spacecraft, as the system designs are often very complex and the constraints can be difficult to characterize. There has been little previous research on the use of MBSE as a design space exploration tool or in support of trade-offs. This paper investigates the potential to use MBSE for design exploration and to understand trade-offs, through the creation of a new toolset including a SysML profile. The tool draws on generative design (allowing automatic guided generation of a multitude of design alternatives) and system optimization to rapidly generate and assess new designs using interactive analysis and visualizations. Techniques such as surrogate modelling, genetic algorithms and robustness measurements will be available in the toolset. The toolset was applied to a design scenario aiming to improve the trade off and design selection process for LEO Earth observation satellites. The upcoming ESA TRUTHS space mission was used as a case study and the design process was recorded and compared to a manual design exploration approach. The toolset was found to reduce the design exploration time by 38% to 96% , allow exploration of more designs in an equivalent time and provide better quantification of the relationships present in the design space, all without drops in selected design quality. For now, the toolset can only perform parameter variation in the design exploration and future work is expected to extend this to higher levels of variability. The study also discusses how the MBSE toolset could be applied to other missions, offering the same advantages to all early phase spacecraft designers.
Original languageEnglish
Title of host publication2023 IEEE Aerospace Conference
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages20
ISBN (Electronic)978-1-6654-9032-0
ISBN (Print)978-1-6654-9033-7
DOIs
Publication statusPublished - 15 May 2023
EventIEEE Aerospace Conference - Montana, United States, Big Sky, United States
Duration: 4 Mar 202311 Mar 2023
https://aeroconf.org/

Publication series

NameIEEE Aerospace Conference Proceedings
Volume2023-March
ISSN (Print)1095-323X

Conference

ConferenceIEEE Aerospace Conference
Country/TerritoryUnited States
CityBig Sky
Period4/03/2311/03/23
Internet address

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Fingerprint

Dive into the research topics of 'Towards Improving the Design Space Exploration Process Using Generative Design With MBSE'. Together they form a unique fingerprint.

Cite this