Document Open Access Logo

Usability of Symbolic Regression for Hybrid System Identification - System Classes and Parameters (Short Paper)

Authors Swantje Plambeck , Maximilian Schmidt , Audine Subias , Louise Travé-Massuyès , Goerschwin Fey

PDF
Thumbnail PDF

File

OASIcs.DX.2024.30.pdf
  • Filesize: 1.21 MB
  • 14 pages

Document Identifiers

Author Details

Swantje Plambeck
  • Institute of Embedded Systems, Hamburg University of Technology, Germany
Maximilian Schmidt
  • Institute of Embedded Systems, Hamburg University of Technology, Germany
Audine Subias
  • LAAS-CNRS, Université de Toulouse, INSA, France
Louise Travé-Massuyès
  • LAAS-CNRS, Université de Toulouse, CNRS, France
Goerschwin Fey
  • Institute of Embedded Systems, Hamburg University of Technology, Germany

Funding

This work was supported by a fellowship of the German Academic Exchange Service (DAAD) and the ECIU Universities. The research is partially funded by the BMBF project AGenC no. 01IS22047A. It is also supported by ANITI through the French "Investing for the Future – P3IA" program under the Grant agreement n^oANR-19-P3IA-0004.

Acknowledgements

Furthermore, we would like to thank Nicola Zaupa and Luca Zaccarian from LAAS CNRS for their support and comments on the power converter example.

Cite As Get BibTex

Swantje Plambeck, Maximilian Schmidt, Audine Subias, Louise Travé-Massuyès, and Goerschwin Fey. Usability of Symbolic Regression for Hybrid System Identification - System Classes and Parameters (Short Paper). In 35th International Conference on Principles of Diagnosis and Resilient Systems (DX 2024). Open Access Series in Informatics (OASIcs), Volume 125, pp. 30:1-30:14, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2024) https://doi.org/10.4230/OASIcs.DX.2024.30

Abstract

Hybrid systems, which combine both continuous and discrete behavior, are used in many fields, including robotics, biological systems, and control systems. However, due to their complexity, finding an accurate model is a challenge. This paper discusses the usage of symbolic regression to learn hybrid systems from data and specifically analyses learning parameters for a recent algorithm. Symbolic regression is a powerful tool that can automatically discover accurate and interpretable mathematical models in the form of symbolic expressions.
Models generated by symbolic regression are a valuable tool for system identification and diagnosis, e.g., to predict future system behavior or detect anomalies. A major opportunity of our approach is the ability to detect transitions between different continuous behaviors of a system directly based on the dynamics. From a diagnosis perspective, this can advantageously be used to detect the system entering fault modes and identify their models. This paper presents a parameter study for a symbolic regression based identification algorithm.

Subject Classification

ACM Subject Classification
  • Computer systems organization → Embedded and cyber-physical systems
  • Computing methodologies → Symbolic and algebraic algorithms
  • Computing methodologies → Learning paradigms
  • Computing methodologies → Modeling methodologies
Keywords
  • Hybrid Systems
  • Symbolic Regression
  • System Identification

Metrics

  • Access Statistics
  • Total Accesses (updated on a weekly basis)
    0
    PDF Downloads
    0
    Metadata Views

Supplementary Materials

References

  1. Takuya Akiba, Shotaro Sano, Toshihiko Yanase, Takeru Ohta, and Masanori Koyama. Optuna: A next-generation hyperparameter optimization framework, 2019. URL: https://arxiv.org/abs/1907.10902.
  2. Rajeev Alur. Principles of Cyber-Physical Systems. Technical report, MIT Press, 2015. Google Scholar
  3. Nathalie Barbosa Roa, Louise Travé-Massuyès, and Victor H. Grisales-Palacio. Dyclee: Dynamic clustering for tracking evolving environments. Pattern Recognition, 94:162-186, 2019. URL: https://doi.org/10.1016/J.PATCOG.2019.05.024.
  4. B. Ould Bouamama, R. Mrani Alaoui, P. Taillibert, and M. Staroswiecki. Diagnosis of a two-tank system. Technical report, Intern Report of CHEM-project, USTL, 2001. Google Scholar
  5. Michael S. Branicky. Introduction to Hybrid Systems, pages 91-116. Birkhäuser, Boston, 2005. Google Scholar
  6. Kevin René Broløs, Meera Vieira Machado, Chris Cave, Jaan Kasak, Valdemar Stentoft-Hansen, Victor Galindo Batanero, Tom Jelen, and Casper Wilstrup. An approach to symbolic regression using feyn. arXiv, 2021. URL: https://arxiv.org/abs/2104.05417.
  7. Frank J. Christophersen. Piecewise Affine Systems, pages 39-42. Springer, Berlin Heidelberg, 2007. Google Scholar
  8. Oswaldo Luiz Valle Costa, Ricardo Paulino Marques, and Marcelo Dutra Fragoso. Markov Jump Linear Systems, pages 1-14. Springer, London, 2005. Google Scholar
  9. Miles Cranmer. Interpretable machine learning for science with pysr and symbolicregression.jl, 2023. https://arxiv.org/abs/2305.01582, URL: https://doi.org/10.48550/arXiv.2305.01582.
  10. Robert Feldt and Peter Nordin. Using factorial experiments to evaluate the effect of genetic programming parameters. In Genetic Programming, 2000. Google Scholar
  11. Louis Goupil, Elodie Chanthery, Louise Travé-Massuyès, and Sébastien Delautier. Tree based diagnosis enhanced with meta knowledge. In International Workshop on Principles of Diagnosis (DX), 2023. Google Scholar
  12. F. Hutter, H. Hoos, and K. Leyton-Brown. An efficient approach for assessing hyperparameter importance. In International Conference on Machine Learning (ICML), 2014. Google Scholar
  13. Hassan K. Khalil. Nonlinear systems. Pearson Education International, 3. edition, 2000. Google Scholar
  14. John R. Koza. Genetic Programming: On the programming of computers by means of natural selection. MIT Press, 1994. Google Scholar
  15. Gabriel Kronberger, Lukas Kammerer, and Michael Kommenda. Identification of dynamical systems using symbolic regression. arXiv, 2021. URL: https://arxiv.org/abs/2107.06131.
  16. William B. Langdon. Genetic programming convergence. In Genetic and Evolutionary Computation Conference Companion, 2022. Google Scholar
  17. Thomas Loveard and Vic Ciesielski. Genetic programming for classification: An analysis of convergence behaviour. Lecture Notes in Artificial Intelligence, 2557:309-320, 2002. URL: https://doi.org/10.1007/3-540-36187-1_27.
  18. Daniel L. Ly and Hod Lipson. Learning symbolic representations of hybrid dynamical systems. Journal of Machine Learning Research, 13(115):3585-3618, 2012. URL: https://doi.org/10.5555/2503308.2503356.
  19. Oliver Niggemann, Benno Stein, Asmir Vodencarevic, Alexander Maier, and Hans Kleine Büning. Learning behavior models for hybrid timed systems. In AAAI Conference on Artificial Intelligence, 2012. Google Scholar
  20. Brenden K Petersen, Mikel Landajuela Larma, Terrell N. Mundhenk, Claudio Prata Santiago, Soo Kyung Kim, and Joanne Taery Kim. Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients. In International Conference on Learning Representations, 2021. Google Scholar
  21. Swantje Plambeck. Symbolic Regression for Hybrid Automata, 2024. Software, (visited on 2024-11-12). URL: https://github.com/TUHH-IES/SymbolicRegression4HA
    Software Heritage Logo archived version
    full metadata available at: https://doi.org/10.4230/artifacts.22520
  22. Swantje Plambeck, Aaron Bracht, Nemanja Hranisavljevic, and Goerschwin Fey. Famos– fast model learning for hybrid cyber-physical systems using decision trees. In International Conference on Hybrid Systems: Computation and Control, 2024. Google Scholar
  23. Swantje Plambeck, Maximilian Schmidt, Audine Subias, Louise Travé-Massuyès, and Goerschwin Fey. Dynamics-based identification of hybrid systems using symbolic regression. In Software Engineering and Advanced Applications (SEAA), 2024. Google Scholar
  24. Riccardo Poli, William B Langdon, and Nicholas F McPhee. A Field Guide to Genetic Programming, volume 10. Springer, 2008. Google Scholar
  25. Iman Saberi, Fathiyeh Faghih, and Farzad Sobhi Bavil. A passive online technique for learning hybrid automata from input/output traces. ACM Transactions on Embedded Computing Systems, 22(1), October 2022. URL: https://doi.org/10.1145/3556543.
  26. Michael Schmidt and Hod Lipson. Distilling free-form natural laws from experimental data. Science, 324(5923):81-85, 2009. Google Scholar
  27. Johan Schoukens and Lennart Ljung. Nonlinear system identification: A user-oriented road map. IEEE Control Systems Magazine, 39(6):28-99, 2019. Google Scholar
  28. E. Sontag. Nonlinear regulation: The piecewise linear approach. IEEE Transactions on Automatic Control, 26(2):346-358, 1981. Google Scholar
  29. Xiaodong Yang, Omar Ali Beg, Matthew Kenigsberg, and Taylor T. Johnson. A framework for identification and validation of affine hybrid automata from input-output traces. ACM Transactions on Cyber-Physical Systems, 6(2):1-24, 2022. URL: https://doi.org/10.1145/3470455.
  30. Nicola Zaupa, Luis Martínez-Salamero, Carlos Olalla, and Luca Zaccarian. Hybrid control of self-oscillating resonant converters. IEEE Transactions on Control Systems Technology, 31(2):881-888, 2023. URL: https://doi.org/10.1109/TCST.2022.3179948.
Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail
© 2023-2024 Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact