Options
FaMoS– Fast Model Learning for Hybrid Cyber-Physical Systems using Decision Trees
Citation Link: https://doi.org/10.15480/882.9569
Publikationstyp
Conference Paper
Date Issued
2024-05-14
Sprache
English
TORE-DOI
Article Number
7
Citation
HSCC '24: Proceedings of the 27th ACM International Conference on Hybrid Systems: Computation and Control. - Art. no. 7 (2024)
Contribution to Conference
Publisher DOI
Scopus ID
Publisher
ACM
ISBN
979-8-4007-0522-9
Peer Reviewed
true
In the domain of cyber-physical systems, there is an increasing relevance of data-driven approaches for the learning of hybrid system dynamics. In particular, accurate models have been successfully abstracted from continuous (real-valued) traces and applied for various goals. However, industrial applications involving online modeling or rapid prototyping have two additional requirements: 1) runtime efficiency and 2) the interpretability of the approach and results.
This work adopts a common break down of this learning problem into four steps: 1) trace segmentation, 2) segment clustering, 3) characterization of the dynamics for each cluster (mode) and 4) learning of the overall model of mode transitions. Correspondingly, the bottlenecks in the state-of-the-art approaches are identified and discussed. Then, in a heuristic manner, interpretable and time-efficient algorithms for each of the steps are proposed giving a novel approach named FaMoS. The accuracy and runtime efficiency of the approach are evaluated for several system examples. FaMoS shows very short learning time, while the model’s predictions of system dynamics are close to the ground truth behavior.
This work adopts a common break down of this learning problem into four steps: 1) trace segmentation, 2) segment clustering, 3) characterization of the dynamics for each cluster (mode) and 4) learning of the overall model of mode transitions. Correspondingly, the bottlenecks in the state-of-the-art approaches are identified and discussed. Then, in a heuristic manner, interpretable and time-efficient algorithms for each of the steps are proposed giving a novel approach named FaMoS. The accuracy and runtime efficiency of the approach are evaluated for several system examples. FaMoS shows very short learning time, while the model’s predictions of system dynamics are close to the ground truth behavior.
Subjects
Cyber-Physical Systems
Decision Trees
Hybrid Dynamical Systems
Machine Learning
MLE@TUHH
DDC Class
004: Computer Sciences
620: Engineering
Funding Organisations
Digitization and Technology Research Center of the Federal Armed Forces of Germany (BMVg)
More Funding Information
This research is partially funded by the two sources BMBF project AGenC no. 01IS22047A and (K)ISS (https://dtecbw.de/home/forschung/hsu/projekt-kiss) as part of dtec.bw® - Digitization and Technology Research Center of the Federal Armed Forces of Germany (BMVg) which we gratefully acknowledge. dtec.bw is funded by the European Union – NextGenerationEU
Publication version
publishedVersion
Loading...
Name
2023_cpsweek.pdf
Type
Main Article
Size
694.42 KB
Format
Adobe PDF