Abstract
Automation in agricultural machinery is a crucial driver of productivity and sustainability. Some automation features like automated steering and real-time data analytics are already state-of-the-art. On the other hand, a human driver performs the optimization of the working speed manually, and the automation of this is an ongoing challenge. Process quality and process efficiency are the two main targets in this optimization. Agricultural soil tillage requires achieving both. Therefore, the correlation between process quality optimization and process efficiency is fundamental, and vice versa. The approach presented in this paper shows how the two optimization targets of efficiency and process quality can be optimized and aligned together. Optical sensors determine various parameters to describe and model the process quality. The measured machine state determines the characteristics of the interaction forces between the machine and the environment. A machine learning algorithm describes the relationships in the drivetrain. The two process targets are each predicted for different working speeds and are combined in the form of a boundary target and an optimization target to identify one optimized target speed value.
Zusammenfassung
Die Automatisierung von Landmaschinen ist ein entscheidender Faktor für Produktivität und Nachhaltigkeit. Einige Automatisierungsfunktionen wie automatische Lenkung und Datenanalyse in Echtzeit sind bereits Stand der Technik. Gleichzeitig führt ein menschlicher Fahrer die Optimierung der Arbeitsgeschwindigkeit weiterhin manuell durch, und die Automatisierung dieses Vorgangs ist bis heute eine Herausforderung. Prozessqualität und Prozesseffizienz sind die beiden Hauptziele bei dieser Optimierung. Bei der landwirtschaftlichen Bodenbearbeitung muss beides zur gleichen Zeit erreicht werden. Daher ist die Korrelation zwischen der Optimierung der Prozessqualität und der Prozesseffizienz von grundlegender Bedeutung und umgekehrt. Der in diesem Beitrag vorgestellte Ansatz zeigt, wie die beiden Optimierungsziele Effizienz und Prozessqualität gemeinsam optimiert und aufeinander abgestimmt werden können. Optische Sensoren ermitteln verschiedene Parameter zur Beschreibung und Modellierung der Prozessqualität. Der gemessene Maschinenzustand bestimmt die Eigenschaften der Wechselwirkungskräfte zwischen Maschine und Umgebung. Ein maschineller Lernalgorithmus beschreibt die Zusammenhänge im Antriebsstrang. Die beiden Prozessziele werden jeweils für unterschiedliche Arbeitsgeschwindigkeiten vorhergesagt und in Form eines Schrankenziels und eines Optimierungsziels zu einem optimierten Sollgeschwindigkeitswert zusammengeführt.
About the authors
M.Sc. Benjamin Kazenwadel received his master's degree in mechanical engineering from Karlsruhe Institute of Technology. In his master's thesis, he already worked on the use of machine learning for energy efficiency optimization in mobile machines. In his research as a research assistant, he continues this work.
M.Sc. Simon Becker has been working as a research assistant at KIT since 2017. His research topics are the use of machine learning for the control of mobile machines and process sensor technology. In the meantime, he works as a chief engineer while continuing to work on these scientific topics.
M.Sc. Marina Graf graduated from KIT in 2021 with a degree in mechanical engineering. She focused on mechatronics and information technology during her master studies. Her main research topic is process sensor technology in agricultural applications.
Prof. Dr.-Ing. Marcus Geimer leads the Institute of Mobile Machines at KIT since 2005. The main research focus is on automation systems, drive train concepts, hydraulics and simulation. Machine learning methods are investigated to control and automate mobile machines in agricultural and forestal applications.
Acknowledgment
We would like to thank AGCO GmbH for supporting the research project.
-
Research ethics: Not applicable.
-
Author contributions: The authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Competing interests: The authors state no conflict of interest.
-
Research funding: None declared.
-
Data availability: Not applicable.
References
[1] O. Bosse and W.-D. Kalk, “Kenngröße zum Bewerten von Bodenbearbeitungswerkzeugen und -geräten bei experimentellen Vergleichen,” Grundlagen Landtechnik, vol. 4, no. 38, pp. 106–113, 1988.Search in Google Scholar
[2] T. Riegler, C. Rechberger, F. Handler, and H. Prankl, “Bildverarbeitungssystem zur Qualitätsbeurteilung von Bodenbearbeitung,” LAND.TECHNIK 2014, vol. 69, no. 3, pp. 125–131, 2014.Search in Google Scholar
[3] S. Steinhaus, Methodik zur Bewertung und Erfassung der Effektivität und Effizienz von landwirtschaftlichen Verfahren und Prozessen: Dissertation, ser. Forschungsberichte aus dem Institut für mobile Maschinen und Nutzfahrzeuge, Düren, Shaker Verlag, 2022.Search in Google Scholar
[4] P. Riegler-Nurscher, J. Karner, J. Huber, et al.., “A system for online control of a rotary harrow using soil roughness detection based on stereo vision,” in LAND.TECHNIK 2017, VDI Verlag, 2017, pp. 559–566.10.51202/9783181023006-559Search in Google Scholar
[5] M. Schmidt, “AI-based tillage job quality assessment for advanced machine automation in agriculture,” in LAND.TECHNIK 2022, Germany, VDI Verlag, 2022, pp. 567–572.10.51202/9783181024065-567Search in Google Scholar
[6] O. Bosse, C. Bernard, H. Petelkau, and A. Kunze, “Vorschläge zur Definition von Begriffen in der Bodenbearbeitung,” agrartechnik, vol. 28, no. 6, pp. 248–249, 1978.Search in Google Scholar
[7] M. Schreiber, “Kraftstoffverbrauch beim Einsatz von Ackerschleppern im besonderen Hinblick auf die CO2-Emissionen,” Ph.D. dissertation, 2012 [Online]. Available at: http://opus.uni-hohenheim.de/volltexte/2012/643/.Search in Google Scholar
[8] B. Li, D. Sun, M. Hu, et al.., “Automatic gear-shifting strategy for fuel saving by tractors based on real-time identification of draught force characteristics,” Biosyst. Eng., vol. 193, pp. 46–61, 2020. https://doi.org/10.1016/j.biosystemseng.2020.02.008.Search in Google Scholar
[9] S. Becker, K. Daiss, K. Daaboul, M. Geimer, and M. J. Zöllner, “Machine learning for process automation of mobile machines in field applications,” in LAND.TECHNIK 2019 : Hannover, Nov. 8th + 9th 2019 : 77th International Conference on Agricultural Engineering, 2019, p. 187 [Online]. Available at: https://publikationen.bibliothek.kit.edu/1000099805.Search in Google Scholar
[10] S. Becker, B. Kazenwadel, and M. Geimer, “Automation and optimization of working speed and depth in agricultural soil tillage with a model predictive control based on machine learning,” in LAND.TECHNIK 2022, VDI Verlag, 2022, pp. 55–64 [Online]. Available at: https://publikationen.bibliothek.kit.edu/1000143315.10.51202/9783181023952-55Search in Google Scholar
[11] M. Peeters, V. Kloster, T. Fedde, and L. Frerichs, “Integrated wheel load measurement for tractors,” in LAND.TECHNIK 2017, VDI Verlag, 2017, pp. 423–430 [Online]. Available at: https://publikationen.bibliothek.kit.edu/1000143315.10.51202/9783181023006-423Search in Google Scholar
[12] F. Zoz and R. Grisso, “Traction and tractor performance,” ASAE Distinguish. Ser., vol. 27, pp. 1–47, 2012.Search in Google Scholar
[13] ASAE S296.5. General Terminology for Traction of Agricultural Traction and Transport Devices and Vehicles, USA, American Society of Agricultural and Biological Engineers (ASABE), 2009.Search in Google Scholar
[14] B. Pichlmaier, “Traktionsmanagement für Traktoren,” Ph.D. dissertation, Technische Universität München, 2012 [Online]. Available at: https://mediatum.ub.tum.de/1098891.Search in Google Scholar
[15] W. Brixius, “Traction prediction equations for bias ply tires,” ASAE, St. Joseph, MI, Tech. Rep. ASAE paper No. 87-1622, 1987.Search in Google Scholar
[16] ASAE D497.7. Agricultural Machinery Management Data, St. Joseph, MI, American Society of Agricultural and Biological Engineers (ASABE), 2011 [Online]. Available at: https://elibrary.asabe.org/abstract.asp?aid=36431&t=2.Search in Google Scholar
[17] K.-T. Renius, Fundamentals of Tractor Design, Cham, Springer International Publishing, 2020.10.1007/978-3-030-32804-7Search in Google Scholar
[18] M. Geimer, Mobile Working Machines, Warrendale, Pennsylvania, USA, SAE International, 2020.10.4271/9780768094329Search in Google Scholar
[19] A. Kobelski, P. Osinenko, and S. Streif, “A method of online traction parameter identification and mapping,” IFAC-PapersOnLine, vol. 53, no. 2, pp. 13 933–13 938, 2020. https://doi.org/10.1016/j.ifacol.2020.12.909.Search in Google Scholar
[20] T. Harrigan and C. Rotz, “Draft relationships for tillage and seeding equipment,” Appl. Eng. Agric., vol. 11, pp. 773–783, 1995. https://doi.org/10.13031/2013.25801.Search in Google Scholar
[21] K. Köller and O. Hensel, Eds. Verfahrenstechnik in der Pflanzenproduktion, Stuttgart, Verlag Eugen Ulmer, 2019.10.36198/9783838551982Search in Google Scholar
[22] “Gütevorschriften für Arbeiten der Pflanzenproduktion,” in Akademie der Landwirtschaftswissenschaften der DDR, Berlin, Tech. Rep., 1984.Search in Google Scholar
© 2023 Walter de Gruyter GmbH, Berlin/Boston
