Autonomno upravljanje bespilotnom letjelicom primjenom konvolucijske neuronske mreže

master's thesis

• Title (english): Autonomous control of an UAV by using convolutional neural network
• Author: Ivan Bralić
• Mentor: Stjepan Bogdan (mentor)
• Committee member: Stjepan Bogdan (predsjednik povjerenstva)
• Committee member: Zdenko Kovačić (član povjerenstva)
• Committee member: Matko Orsag (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-16, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Tema rada je bila ostvariti autonomno upravljanje bespilotnom letjelicom koristeći metode dubokog učenja. U radu je istraženo nekoliko mogućnosti efikasnih arhitektura konvolucijskih neuronskih mrežu u svrhe semantičke segmentacije. Koristeći informacije o lokaciji ceste sa segmentirane slike ostvarena je navigacija letjelicom za praćenje iste. Implementirani algoritam upravljanja je kao referentne veličine zadavao brzinu gibanja i potreban zakret letjelice u svrhu uspješnog praćenja prometnice. Eksperimentalnom analizom je potvrđena valjanost predloženih metoda i ostvareni su zadovoljavajući rezultati kvalitete detekcije i navigacije u simulacijskom okruženju.

Abstract (english)

The goal of this thesis was to accomplish autonomous control of unmanned aerial vehicle by using methods of deep learning. Several architectures of deep convolutional neural networks were explored for the purpose of semantic segmentation. Based on the information provided by the segmented image, autonomous navigation algorithm was developed for the goal of road following. Implemented control algorithm generated flying velocity and needed yaw angle in order for UAV to successfully follow the road. Validity of methods were experimentally analyzed. Results from the simulation environments showed satisfactory results of detection and navigation.

• Keywords: semantička segmentacija; duboko učenje; bespilotne letjelice; autonomna navigacija; praćenje ceste
• Keywords (english): semantic segmentation; deep learning; UAV; autonomous navigation; road following
• Language: croatian
• URN:NBN: urn:nbn:hr:168:145832
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 11:52:20