A multi-objective evolutionary approach to discover explainability trade-offs when using linear regression to effectively model the dynamic thermal behaviour of electrical machines.

Banda, Tiwonge Msulira; Zăvoianu, Alexandru-Ciprian; Petrovski, Andrei; Wöckinger, Daniel; Bramerdorfer, Gerd

Abstract

Modelling and controlling heat transfer in rotating electrical machines is very important as it enables the design of assemblies (e.g., motors) that are efficient and durable under multiple operational scenarios. To address the challenge of deriving accurate data-driven estimators of key motor temperatures, we propose a multi-objective strategy for creating Linear Regression (LR) models that integrate optimised synthetic features. The main strength of our approach is that it provides decision makers with a clear overview of the optimal trade-offs between data collection costs, the expected modelling errors and the overall explainability of the generated thermal models. Moreover, as parsimonious models are required for both microcontroller deployment and domain expert interpretation, our modelling strategy contains a simple but effective step-wise regularisation technique that can be applied to outline domain-relevant mappings between LR variables and thermal profiling capabilities. Results indicate that our approach can generate accurate LR-based dynamic thermal models when training on data associated with a limited set of load points within the safe operating area of the electrical machine under study.

Citation

BANDA, T.M., ZĂVOIANU, A.-C., PETROVSKI, A., WÖCKINGER, D. and BRAMERDORFER, G. 2024. A multi-objective evolutionary approach to discover explainability trade-offs when using linear regression to effectively model the dynamic thermal behaviour of electrical machines. ACM transactions on evolutionary learning and optimization [online], 4(1), article number 3. Available from: https://doi.org/10.1145/3597618

• Journal Article Type: Article
• Acceptance Date: Apr 27, 2023
• Online Publication Date: May 19, 2023
• Publication Date: Mar 31, 2024
• Deposit Date: Sep 22, 2023
• Publicly Available Date: Oct 10, 2023
• Journal: ACM transactions on evolutionary learning and optimization
• Print ISSN: 2688-299X
• Electronic ISSN: 2688-3007
• Publisher: Association for Computing Machinery (ACM)
• Peer Reviewed: Peer Reviewed
• Volume: 4
• Issue: 1
• Article Number: 3
• DOI: https://doi.org/10.1145/3597618
• Keywords: Data-driven models; Electrical machines; Linear regression; Explainability; Problem formalisation; Cost; Accuracy; Computing methodologies; Optimization algorithms; Theory of computation; Evolutionary algorithms
• Public URL: https://rgu-repository.worktribe.com/output/2086616