Lokalizacija vozila zasnovana na fuziji podataka višesenzorskog sustava

master's thesis

• Title (english): Vehicle localization based on multi-sensor system data fusion
• Author: Fran Raos
• Mentor: Marija Seder (mentor)
• Committee member: Marija Seder (predsjednik povjerenstva)
• Committee member: Ivan Petrović (član povjerenstva)
• Committee member: Ivan Marković (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-10, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Ovaj diplomski rad sastoji se od četiri poglavlja te uvoda, zaključka i priloga. U drugom poglavlju obrađuje se matematički model odabranog algoritma lokalizacije prema odabranom rješenju te algoritam neizravnog proširenog Kalmanova filtra koji se koristi u procesu fuzije podataka senzora. Mjerenja senzora koja se fuziraju su: obrađena mjerenja GNSS-a, mjerenja vizualne odometrije dobivena kamerom, simulirana mjerenja enkodera te neobrađena mjerenja IMU senzora. U trećem poglavlju obrađuje se vlastita implementacija algoritma lokalizacije u MATLAB-u prema prethodno objašnjenjom matematičkom modelu. Navode se sve funkcionalnosti napravljenih i korištenih skripti. Sve skripte prilažu se uz ovaj diplomski rad u elektroničkom obliku. U četvrtom poglavlju nalaze se rezultati estimacije algoritma lokalizacije pri čemu su za testiranje algoritma estimacije korištena tri KITTI dataseta. Uz rezultate nalaze se i komentari dobivenih rezultata estimacije. U zadnjem poglavlju nalaze se prijedlozi za budući rad na algoritmu. Cilj ovoga diplomskoga rada bio je implementirati algoritam lokalizacije prema unaprijed odabranoj literaturi te ostvariti uspješnu estimaciju lokalizacije. Ciljevi ovog diplomskog rada su zadovoljeni.

Abstract (english)

This master thesis consists of four chapters, introduction, conclusion and appendix. The mathematical model of the chosen localization algorithm and the indirect extended Kalman filter algorithm which is used in the data fusion process are explained in the second chapter. Multi-sensor data in the fusion process are: processed GNSS measurements, visual odometry measurements from the camera, simulated encoder measurements and raw IMU measurements. The own implementation of the localization algorithm in MATLAB based on the mentioned mathematical model is explained in the third chapter. All functionalities of the made and used scripts are listed. All scripts come with this master thesis in the electronic form. The results of the implemented localization algorithm which was tested on three KITTI datasets is presented in the fourth chapter. The comments on the results are also noted. In the last chapter the suggestions on future work are listed. The goal of this master thesis was to implement the localization algorithm based on previously selected literature and to make a successful localization estimation. The master thesis goals were satisfied.

• Keywords: EKF; ES-EKF; neizravni prošireni Kalmanov filtar; lokalizacija; algoritam fuzije; estimacija pogreške; GNSS; obrađeni GNSS; vizualna odometrija; VO; MATLAB; implementacija; KITTI
• Keywords (english): EKF; ES-EKF; indirect extended Kalman filter; localization; fusion algorithm; error state estimation; GNSS; processed GNSS; GNSS loose coupling; visual odometry; VO; MATALAB; implementation; KITTI
• Language: croatian
• URN:NBN: urn:nbn:hr:168:373281
• 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 13:40:47