Gaussian Process Regression as a precrash velocity determination method–subcompact vehicle class

• Autorzy: Turoboś Filip , Mrowicki Adam , Konieczny Piotr , Madziara Szymon , Stajuda Łukasz , Šarkan Branislav , Kubiak Przemysław , Levchenko Dymytro , Meisner Nicole

Identyfikatory

DOI

10.14669/AM/189172

• Języki publikacji: EN

Abstrakty

EN

The following paper presents an innovative approach to determining vehicle precrash velocity when hitting an immovable obstacle facing forward. Precrash velocity is necessary in order to perform a crash reconstruction. It is needed for the time-space analysis of the events, as well as to assess crash mitigation and to evaluate drivers’ technique and tactics. For this task, the authors are using Gaussian Process Regression (GPR). Such an approach offers a number of advantages over the currently used methods that prove to be outdated when considering modern vehicles. The mathematical model was trained on a database shared by the National Highway Traffic Safety Administration. This database covers a large number of crash tests of different kind, however authors focus on frontal collisions of the subcompact car class. Due to low accuracy of linear methods used up till now, Authors developed an innovative approach to determine the EES parameter utilizing Gaussian process regression. The newly developed method is an effective and accurate way to determine the vehicle’s velocity and shows promising results, as is demonstrated in this paper.

Słowa kluczowe

EN

car crash reconstruction; EES; Gaussian Process Regression

• Wydawca: Wydawnictwo Naukowe Sieci Badawczej Łukasiewicz - Przemysłowego Instytutu Motoryzacji
• Czasopismo: Archiwum Motoryzacji
• Rocznik: 2024
• Tom: Vol. 105, nr 3
• Strony: 65--73
• Opis fizyczny: Bibliogr. 11 poz., rys.

Twórcy

autor

Turoboś Filip

• Institute of Mathematics, Lodz University of Technology, Poland

• https://orcid.org/0000-0002-5782-6159

autor

Mrowicki Adam

• Institute of Vehicles and Construction Machinery Engineering, Warsaw University of Technology, Poland

• https://orcid.org/0000-0002-7865-099X

autor

Konieczny Piotr

• Institute of Mathematics, Lodz University of Technology, Poland

• https://orcid.org/0000-0003-4755-744X

autor

Madziara Szymon

• Institute of Vehicles and Construction Machinery Engineering, Warsaw University of Technology, Poland

• https://orcid.org/0000-0002-6193-9260

autor

Stajuda Łukasz

• Division of Ecotechnics, Lodz University of Technology, Poland

• https://orcid.org/0009-0008-5086-9245

autor

Šarkan Branislav

• Faculty of Operation and Economics of Transport and Communications, University of Žilina, Slovak Republic

• https://orcid.org/0000-0002-5036-9223

autor

Kubiak Przemysław

• Division of Ecotechnics, Lodz University of Technology, Poland

• https://orcid.org/0000-0003-0225-9609

autor

Levchenko Dymytro

• Division of Ecotechnics, Lodz University of Technology, Poland

• https://orcid.org/0000-0003-0766-4371

autor

Meisner Nicole

• Department of Mathematics, University of Warsaw, Poland

• https://orcid.org/0009-0006-3798-3196

Bibliografia

• [1] Aleksandrowicz P, Aleksandrowicz I, Kukiełka K, Patyk R, Stanowski P. Problem with determining the vehicle impact velocity for car bodies breaking apart. Transport Problems. 2022;17(3):75–86. https://doi.org/10.20858/tp.2022.17.3.07.
• [2] Dudziak A, Caban J, Stopka O, Stoma M, Sejkorová M, Stopková M. Vehicle Market Analysis of Drivers’ Preferences in Terms of the Propulsion Systems: The Czech Case Study. Energies. 2023;16(5):2418. https://doi.org/10.3390/en16052418.
• [3] Flaxman S, Gelman A, Neill D, Smola A, Vehtari A, Wilson AG. Fast hierarchical Gaussian processes. 2015.
• [4] Gershman SJ, Blei DM. A tutorial on Bayesian nonparametric models. Journal of Mathematical Psychology. 2023;56(1):1–12. https://doi.org/10.1016/j.jmp.2011.08.004.
• [5] Guan J, Yao Y, Zhao W, Hagiwara I, Zhao X. Development of an Impact Energy Absorption Structure by an Arc Shape Stroke Origami Type Hydraulic Damper. Shock and Vibration. 2023;2023:1–11.https://doi.org/10.1155/2023/4578613.
• [6] Guzek M, Jackowski J, Jurecki RS, Szumska EM, Zdanowicz P, Żmuda M. Electric Vehicles—An Overview of Current Issues—Part 2—Infrastructure and Road Safety. Energies. 2024;17(2):495. https://doi.org/10.3390/en17020495.
• [7] Liashchynskyi P, Liashchynskyi P. Grid search, random search, genetic algorithm: a big comparison for NAS. preprint. https://doi.org/10.48550/arXiv.1912.06059.
• [8] Moravcová P, Bucsuházy K, Zůvala R, Semela M, Bradáč A. What should I use to calculate vehicle EES?. Plos one. 2024;19(2):e0297940. https://doi.org/10.1371/journal.pone.0297940.
• [9] Williams CK, Rasmussen CE. Gaussian processes for machine learning Cambridge. MA: MIT Press. 2006.
• [10] Williams CK. Prediction with Gaussian processes: From linear regression to linear prediction and beyond. Learning in graphical models. Dordrecht: Springer Netherlands. 1998;89:599–621. https://doi.org/10.1007/978-94-011-5014-9_23.
• [11] Zou T, Liu Y, Zhang Y. A method for analyzing accident reconstruction results under complex uncertain conditions. International journal of crashworthiness. 2023;28(2):224–234. https://doi.org/10.1080/13588265.2022.2074721.