Back to search

Democritus University of Thrace (DUTH)

Τεχνικές ελέγχου και διαχείρισης ενέργειας σε ηλεκτρικά οχήματα που φορτίζονται από το δίκτυο

Abstract

dc:description

Technologies are changing and evolving. The energy sector has been undergoing a global transformation in recent years. One of the rapidly evolving sectors is electric vehicles (EVs). Electric vehicles have become widespread over the last decade because of their advantages over conventional vehicles. However, the widespread penetration of electric vehicles in (global) transport has led to increased electricity demands, which is going to affect the operation of energy distribution systems. This thesis contributes to the smooth integration of electric vehicles into the networks, whose operation is expected to be significantly and negatively affected, according to literature. At the same time, the integration of renewable energy plants is also considered as equally important technology, being at the center of this global transformation. Therefore, in this thesis, a study of the impact of uncontrolled charging of electric vehicles in low-voltage networks is carried out, and a fuzzy energy management strategy to coordinate the charging of electric vehicles in low-voltage networks, considering for the first time the distance of electric vehicles from transformers in fuzzy management systems, is presented. In particular, the developed system takes the network architecture, the thermal limits of the lines, and the voltages at the yokes into account. In addition, electric vehicles are studied as agent-based models, which are characterized by the model of each electric vehicle, the state of charge of their batteries and the charging power. In particular, the effects of uncontrolled charging are investigated, in which two approaches are considered. The first approach considers the maximum number of electric vehicles that can be charged on the case study network and how this is affected by the loads of its households. The second approach looks at the number of grid hypotheses and line thermal limit exceedances in a set of simulation scenarios. The results of the first approach show that the distance of electric vehicles from the grid substation significantly affects the maximum number of electric vehicles that can be charged on the grid. Based on the results observed in the two approaches, a fuzzy management system is designed to coordinate the charging of electric vehicles, which takes the distance of electric vehicle charging points from the power substation, the charging status of electric vehicle batteries, and the charging delay time of electric vehicles into account. The expected significant increase in global electricity demand, followed by the adoption of renewable energy sources, is leading to the modernization of energy distribution networks, for example through the development of microgrids which can improve the reliability of the electricity system. At the same time, the increased integration of electric vehicles is expected to negatively affect power quality, smooth operation, and investment costs of microgrids. Following the investigation of the problems of interconnected systems, a microgrid topology for the autonomous operation of a low-voltage distribution network is studied. The development of such power management systems is essential to ensure the safe and reliable operation of autonomous microgrids. For this reason, a decentralized energy management system based on multi-agent systems was developed for efficient charging of electric vehicles, now extending the state-of-the-art energy management strategy, which uses a fuzzy logic controller combined with a charging power controller based on fuzzy cognitive maps. According to the literature investigation results, this is the first time that fuzzy cognitive map theory is introduced to electric vehicle charging management problems. The behavior of the proposed decentralized energy management system with multiple agents brought about a significant reduction in the microgrid investment cost, compared to the use of a fuzzy logic controller-based energy management strategy. The total cost of the microgrid for a 20-year investment is reduced by about 8.8%. Moreover, the integration of the fuzzy cognitive map-based charging power controller leads to a significant number of charged electric vehicles. The average number of chargeable electric vehicles increased by 31%. Finally, the proposed power management scheme reduces peak load and load fluctuations by about 17% and 29% respectively, without shifting and delaying electric vehicle charging. Therefore, the proposed novel charging management system offers a smart approach for autonomous distribution networks, considering the high penetration of electric vehicles, showing operational and economic benefits. However, as the existing electricity distribution networks are mostly interconnected, a major challenge that residential distribution networks are going to face with the integration of photovoltaic panels and electric vehicles is the non-synchronized power generation from residential photovoltaic (PV) systems with the consumption profiles of electric vehicles. The necessity of energy storage systems is therefore more than obvious. Energy storage systems have significantly high construction costs, which means that the purchase costs for domestic users can be prohibitive. Therefore, a strategy for optimizing the sizing of small distributed photovoltaic energy storage systems is designed and proposed, taking the specific energy needs of each residence and their electric vehicles into account. The strategy is formulated based on household demand and electric vehicle charging. By applying the theory of fuzzy cognitive maps, a structure is designed to define the correlation between individual energy parameters, with an emphasis on electric vehicles and renewable energy characteristics. Optimization results show that the adopted hybrid approach can significantly reduce energy costs, up to almost 40%, while allowing distribution system operators to integrate additional loads without the need to expand the grid. Finally, based on the extracted results, a brief discussion on the concept of charging electric vehicles from residential RES is presented. Hybrid electricity distribution networks with electric vehicles, residential PV systems and dispersed batteries along the networks are considered. In this approach, a methodology for optimal sizing of PV systems and batteries is designed, in addition to optimal systems for managing the charging power of electric vehicles and system batteries, through the use of multiagent systems theory and fuzzy cognitive maps. Six objective functions are considered simultaneously, through the use of the NSGA-II multi-objective optimization algorithm. The objective of this methodology is to examine the behavior of near-optimal Pareto boundaries for different electric vehicle integration rates in two different solar radiation environments. Conclusions are drawn about the near-optimal Pareto boundaries calculated in the 8 different scenarios considered.

Degree

thesis:*
Grantor dc:publisher
Democritus University of Thrace (DUTH)
Year dc:date
2023

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Boglou, Vasileios
  • Μπόγλου, Βασίλειος

Subjects

dc:subject × 16

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/53692

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

Boglou, Vasileios; Μπόγλου, Βασίλειος. Τεχνικές ελέγχου και διαχείρισης ενέργειας σε ηλεκτρικά οχήματα που φορτίζονται από το δίκτυο. Democritus University of Thrace (DUTH), 2023. http://hdl.handle.net/10442/hedi/53692