prikaz prve stranice dokumenta Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje
  Download
PDF 3.72 MB
master's thesis
Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje
2022. urn:nbn:hr:200:233902
Dumančić, David
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Communications
Chair of Multimedia Systems and Digital Television

Cite this document

Dumančić, D. (2022). Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:233902

Dumančić, David. "Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2022. https://urn.nsk.hr/urn:nbn:hr:200:233902

Dumančić, David. "Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2022. https://urn.nsk.hr/urn:nbn:hr:200:233902

Dumančić, D. (2022). 'Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 01 May 2024, https://urn.nsk.hr/urn:nbn:hr:200:233902

Dumančić D. Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2022 [cited 2024 May 01] Available at: https://urn.nsk.hr/urn:nbn:hr:200:233902

D. Dumančić, "Usporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2022. Available at: https://urn.nsk.hr/urn:nbn:hr:200:233902

Please login to the repository to save this object to your list.
Metadata
TitleUsporedba rada algoritma za autonomnu vožnju zasnovanog na procjeni kuta zakreta upravljača vozila u različitim simulatorima za razvoj algoritama autonomne vožnje
Title (english)Comparison of the performance of autonomous driving algorithm based on vehicle steering angle prediction in different simulators for autonomous driving algorithms development
AuthorDavid Dumančić
MentorMario Vranješ (mentor)
Committee memberMario Vranješ (predsjednik povjerenstva)
Committee memberMarijan Herceg (član povjerenstva)
Committee memberRatko Grbić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Communications)
(Chair of Multimedia Systems and Digital Television)
Osijek
Defense date and country2022-09-27, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Artificial Intelligence
Abstract
U ovom diplomskom radu opisan je problem procjene kuta zakreta upravljača vozila, što uključuje pregled postojećih rješenja za procjenu kuta zakreta upravljača vozila zasnovanih na računalnom vidu, dubokom učenju i podržanom učenju. Dan je osvrt na postojeće simulatore koji omogućuju razvoj algoritama autonomne vožnje. Uspoređeni su CARLA i AirSim simulatori koji su odabrani kao testno okruženje algoritma za autonomnu vožnju. Opisana je implementacija odabranog PilotNet algoritma u simulatore, s ciljem usporedbe mogućnosti koje pružaju navedeni simulatori. Odabrani algoritam procjenjuje kut zakreta upravljača vozila na temelju slike s vozila koju pružaju simulatori, a zatim se šalje procijenjeni kut zakreta upravljača vozila u obliku upravljačke naredbe vozila. Koristeći prednosti koje pružaju simulatori, algoritam je testiran u stvarnom vremenu na simuliranim cestama. Algoritam je testiran na ukupno 198 km vožnje u dvama različitim simulatorima, a pri tome je model 91.65 % vremena uspješno upravljao vozilom u CARLA simulatoru, dok je 78.87 % vremena ispravno upravljao vozilom u AirSim simulatoru. Ispitane su različite pogodnosti koje navedeni simulatori pružaju za razvoj i testiranje algoritama autonomne vožnje. CARLA simulator se pokazao pogodniji za testiranje i razvoj algoritama autonomne vožnje koji se zasnivaju na obradi slike. Sadrži realnije postavke vizualne scene i sadrži veći broj različitih vremenskih uvjeta
Abstract (english)
This graduate thesis describes the problem of vehicle steering angle prediction, which includes a review of existing solutions for vehicle steering angle prediction based on computer vision, deep learning and Reinforcement learning. An overview of existing simulators that enable the development of autonomous driving algorithms is given. CARLA and AirSim simulators, which were selected as the test environment of the autonomous driving algorithm, were compared. The implementation of the selected PilotNet algorithm in simulators is described, with the aim of comparing the possibilities provided by the mentioned simulators. The selected algorithm predicts the steering angle of the vehicle based on the image from the vehicle, provided by the simulators, and then the predicted steering angle of the vehicle is sent in the form of a vehicle control command. Taking advantage of simulators, the algorithm was tested in real time on simulated roads. The algorithm was tested on a total of 198 km of driving in two different simulators, and the model successfully controlled the vehicle 91.65% of the time in the CARLA simulator, while it controlled the vehicle correctly in the AirSim simulator 78.87% of the time. The various benefits that the mentioned simulators provide for the development and testing of autonomous driving algorithms have been examined. The CARLA simulator proved to be more suitable for testing and developing autonomous driving algorithms based on image processing. It contains more realistic visual scene settings and contains a greater number of different weather conditions.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:233902
Study programmeTitle: UNIVERSITY STUDY AUTOMOBILE COMPUTER AND COMMUNICATION
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka automobilskog računarstva i komunikacija (magistar/magistra inženjer/inženjerka automobilskog računarstva i komunikacija)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2022-10-19 09:14:19