Abstract
A methodology to generate excitation signals that can be used for the identification of nonlinear dynamic systems is proposed. In contrast to traditional approaches which are based on specific signal types, the objective here is the homogeneous distribution of the data points in the input space of the model. A space-filling data distribution in the whole input space is a necessity for gathering information about the nonlinearities of the system and minimizes the risk of extrapolation. The methodology can be extended to multiple inputs with moderate increase in complexity which is a key feature for most real-world applications. The quality of the excitation signal is demonstrated with simulations and on a high pressure fuel supply system.
Zusammenfassung
Es wird eine Methode vorgestellt, um Eingangssignale für die Identifikation nichtlinearer dynamischer Systeme zu entwerfen. Im Gegensatz zu Verfahren basierend auf speziellen Signaltypen ist das Ziel hier die gleichmäßige Datenverteilung im Eingangsraum des Modells. Eine raumfüllende Datenverteilung ist notwendig, um alle nichtlinear dynamischen Effekte zu erfassen und das Risiko von Extrapolation zu minimieren. Die hier vorgestellte Strategie lässt sich mit moderatem Aufwand auf mehrere Eingangsgrößen erweitern, was eine Schlüsseleigenschaft für die meisten realen Anwendung darstellt. Die Qualität des optimierten Eingangssignals wird anhand von Simulationen und eines Hochdruck-Kraftstoffversorgungssystems demonstriert.
About the authors
Tim Oliver Heinz started under the supervision of Prof. Nelles as a PhD student at the University of Siegen in 2014 after receiving his master of science in mechanical engineering. His main research topics concern design of dynamic experiments and identification of nonlinear dynamic systems.
Oliver Nelles is Professor at the University of Siegen in the Department of Mechanical Engineering and chair of Automatic Control – Mechatronics. He received his doctor’s degree in 1999 at the Technical University of Darmstadt. His key research topics are nonlinear system identification, dynamics representations, design of experiments, metamodeling, local model networks.
Acknowledgment
The authors would like to thank Mark Schillinger and Benjamin Hartmann for the support of this by providing the measured data of the high pressure fuel supply system.
References
1. Wolf Baumann, Steffen Schaum, Karsten Roepke and Mirko Knaak, Excitation signals for nonlinear dynamic modeling of combustion engines, in: Proceedings of the 17th World Congress, The International Federation of Automatic Control, Seoul, Korea, 2008.Search in Google Scholar
2. Liu Bo, Zhao Jun and Qian Jixin, Design and analysis of test signals for system identification, Computational Science – ICCS 2006, Springer, 2006, pp. 593–600.10.1007/11758532_78Search in Google Scholar
3. George EP Box and R Daniel Meyer, An analysis for unreplicated fractional factorials, Technometrics 28 (1986), 11–18.10.1080/00401706.1986.10488093Search in Google Scholar
4. Marco Cavazzuti, Optimization methods: from theory to design scientific and technological aspects in mechanics, Springer Science & Business Media, 2012.10.1007/978-3-642-31187-1Search in Google Scholar
5. Tobias Ebert, Torsten Fischer, Julian Belz, Tim Oliver Heinz, Geritt Kampmann and Oliver Nelles, Extended Deterministic Local Search Algorithm for Maximin Latin Hypercube Designs, in: Computational Intelligence, 2015 IEEE Symposium Series on, IEEE, pp. 375–382, 2015.10.1109/SSCI.2015.63Search in Google Scholar
6. Matthias Gringard and Andreas Kroll, Zum Optimalen Offline Testsignalentwurf für die Identifikation dynamischer TS-Modelle: Multistufensignale für unsicherheitsminimierte Konklusionsparameter, in: 27. Workshop Computational Intelligence, KIT Scientific Publishing, Dortmund, 2017.Search in Google Scholar
7. Christoph Hametner, Christian Mayr and Stefan Jakubek, Dynamic NOx emission modelling using local model networks, International Journal of Engine Research (2014), 928–933.10.1177/1468087414523281Search in Google Scholar
8. Benjamin Hartmann and Oliver Nelles, Adaptive Test Planning for the Calibration of Combustion Engines-Methodology, Design of experiments (DoE) in engine development (2013), 1–16.Search in Google Scholar
9. Tim Oliver Heinz and Oliver Nelles, Vergleich von Anregungssignalen für Nichtlineare Identifikationsaufgaben, in: Proceedings 26. Workshop Computational Intelligence (F Hoffman, E Hüllermeier and R Mikut, eds.), pp. 139–158, KIT Scientific Publishing, November 2016.Search in Google Scholar
10. Tim Oliver Heinz and Oliver Nelles, Iterative Excitation Signal Design for Nonlinear Dynamic Black-Box Models, Procedia Computer Science (2017), 1054–1061.10.1016/j.procs.2017.08.112Search in Google Scholar
11. Tim Oliver Heinz, Mark Schillinger, Benjamin Hartmann and Oliver Nelles, Excitation Signal Design for Nonlinear Dynamic Systems, in: International Calibration Conference – Automotive Data Analytics, Methods, DoE (Karsten Röpke and Clemens Gühmann, eds.), pp. 191–208, expertVerlag, May 2017.Search in Google Scholar
12. Mark E Johnson, Leslie M Moore and Donald Ylvisaker, Minimax and maximin distance designs, Journal of statistical planning and inference 26 (1990), 131–148.10.1016/0378-3758(90)90122-BSearch in Google Scholar
13. Andreas Kroll, Zum optimalen Testsignalentwurf für die Partitionierung und Teilmodellparametrierung dynamischer Takagi-Sugeno-Modelle: Problemstellung und Lösungsansätze, in: 26. Workshop Computational Intelligence, pp. 97–118, KIT Scientific Publishing, Dortmund, 2016.10.1515/9783110401776Search in Google Scholar
14. Max D Morris and Toby J Mitchell, Exploratory designs for computational experiments, Journal of statistical planning and inference 43 (1995), 381–402.10.1016/0378-3758(94)00035-TSearch in Google Scholar
15. Oliver Nelles, Axes-oblique partitioning strategies for local model networks, in: 2006 IEEE Conference on Computer Aided Control System Design, 2006 IEEE International Conference on Control Applications, 2006 IEEE International Symposium on Intelligent Control, IEEE, pp. 2378–2383, 2006.10.1109/CACSD-CCA-ISIC.2006.4777012Search in Google Scholar
16. Oliver Nelles and Rolf Isermann, Identification of nonlinear dynamic systems classical methods versus radial basis function networks, in: American Control Conference, Proceedings of the 1995, 5, IEEE, pp. 3786–3790, 1995.10.1109/ACC.1995.533847Search in Google Scholar
17. Rik Pintelon and Johan Schoukens, System identification: a frequency domain approach, John Wiley & Sons, 2012.10.1002/9781118287422Search in Google Scholar
18. Luc Pronzato and Werner G Müller, Design of computer experiments: space filling and beyond, Statistics and Computing 22 (2012), 681–701.10.1007/s11222-011-9242-3Search in Google Scholar
19. Nils Tietze, Model-based Calibration of Engine Control Units Using Gaussian Process Regression, (2015).Search in Google Scholar
© 2018 Walter de Gruyter GmbH, Berlin/Boston
