Back to search

National Technical University of Athens (NTUA)

Ευφυείς αλγόριθμοι μάθησης και δρομολόγησης οχημάτων βάσει συνεργατικών σχημάτων

Abstract

dc:description

The aim of the present thesis is to identify the most energy-efficient route towards the desired destination through the implementation of both computational intelligence and communications technologies. In particular, it proposes the implementation of machine learning techniques for reliably estimating the energy cost of the probable paths towards the desired destination based on the previously collected experience. Moreover, in order to further enhance the experience accumulating process, it suggests the use of the available V2V and V2I technologies and the establishment of cooperative networks enabling, thus, the exchange of experience between vehicles. Initially, the thesis identifies the current status and the expected development in the transport field and reveals the size of the energy problem faced globally. In order to limit the expansion of this problem, car manufacturers started developing the so called “green” vehicles (e.g. the fully electric vehicles) that adopt environmentally-friendly technologies. Nevertheless, these green technologies impose some functional limitations to the vehicle’s usage (e.g. in case of fully electric vehicles such limitations are the limited range and the time consuming recharging process). In this context, the objective of the present thesis is to expand the range of fully electric vehicles through the implementation of technologies that belong to fields not related to vehicular development, i.e. the computational intelligence field and the communications field. Not to mention, reviewing the state of the art related to the scope of this thesis renders the adopted approach innovative and rather challenging.Thus, the rest of the thesis is devoted to the development and the verification of the proposed energy-efficient routing algorithms. The development process starts with the formulation of the proposed solution and the justification of the introduced design choices. Forecasting reliably the energy costs of the probable paths towards the destination and confronting with the problems that accrue due to the specificities of the fully electric vehicles (e.g. energy recovery through regenerative breaking) constitute some of the major issues addressed by the present thesis. In particular, in order to address them successfully the thesis proposes the estimation of the route’s energy cost by means of machine learning techniques and promotes the exploitation of travelling experience that is either gathered by the vehicle itself or retrieved from other vehicles through synchronization. Furthermore, the thesis applies suitable shifting techniques so as to eliminate the negative energy costs from routing calculations and develops a methodology for protecting the user’s privacy during the exchange of experience between vehicles.Secondly, the thesis develops the architecture of an intelligent navigation system implementing the introduced routing algorithms. Based on the developed architecture, it further performs a scalability and performance analysis of the designed system. According to the results of this analysis, it is verified that the commercially available hardware and software are capable of supporting the functionality of the developed architecture. Finally, the thesis describes the development of a prototype system implementing the architecture designed previously and it presents the installation of the developed prototype in a fully electric vehicle serving as the test vehicle. This test vehicle performs a series of field trials especially planned for evaluating the efficiency of the proposed energy-efficient routing algorithms and the perceived user experience. The final conclusions of the thesis are extracted based on the generated evaluation report.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2013

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Masikos, Michail
  • Μασίκος, Μιχαήλ

Subjects

dc:subject × 32

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/39412

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Masikos, Michail; Μασίκος, Μιχαήλ. Ευφυείς αλγόριθμοι μάθησης και δρομολόγησης οχημάτων βάσει συνεργατικών σχημάτων. National Technical University of Athens (NTUA), 2013. http://hdl.handle.net/10442/hedi/39412