Back to search

University of Thessaly (UTH)

Platforms for (self) evolving smart energy systems

Abstract

dc:description

The modern society is strongly dependent on electricity which means that the constant supply of such a commodity is essential. On the contrary, while the energy needs of the end-users keep increasing very fast, the expansion of the electricity infrastructure, either talking about the power grid and the transmission system or the power plants themselves does not follow with the same pace that means that the power grid is constantly congested.The advent of Renewable Energy sources (RES) mitigates the supply issues. In contrast, these sources add unpredictability and stochasticity on the grid since nobody can guarantee their uninterrupted contribution in power supply. Moreover, conventional power units have an inherent inertia. They cannot reduce or increase their production and supply instantaneously that means that these units must supply enough amounts that can cover a temporary loss of RES units. Obviously, the extra amounts of power produced by the conventional power plants are wasted if RES operate normally. Electricity cannot be stored. At least not efficiently. In that sense, the extra amounts of the produced power are simply wasted.We should also take into serious consideration the introduction of Electric Vehicles (EVs) which adds more stress on a grid that is already operating at its limits. EVs add an extra load on the system and moreover this load is not diversified too much during the day. More or less, EVs charging cycle follows the same pattern. People leave their homes at morning, go to their works at the same time and charge their cars in the parking lot. They leave their work and return back to their homes where they charge their cars during the afternoon or night.The stability of the power grid is responsibility of the system operators. They are responsible of the load management in order to guarantee the uninterrupted operation of the system. So far, this task is met empirically meaning that system operators have a knowledge about the days or the times that the grid is most congested or stressed and make decisions about how many and which units to dispatch in order to meet the demand and guarantee the safety of the system at the same time.The electricity market is undergoing a transformation from its vertically integrated nature to a competitive one based on a security-constrained, bid-based auctiong system for selling and buying electricity. In this concept that introduced by William W. Hogan, the aforementioned practices become obsolete. Moreover, since producers and consumers compete each other in order to sell/buy energy in the best possible price, Artificial Intelligence (AI) algorithms and methods can offer a solution that maximizes their portfolio on the one hand and on the other hand guarantees the stability of the system in real time.The subject of the current dissertation is on the identification, exploitation and development of novel methods and algorithms focused on the transformation and restructuring of the vertically integrated electricity market into a competitive one where power producers, wholesalers, retailers and end-users compete to optimize their portfolio. A transformation which is of course unavoidable, considering the fossil fuel depletion, and already in progress in the most of the developed countries and especially in the US. To that end, through extensive simulation and experimentation I tried to investigate efficient, non discriminatory ways to exploit electricity auctioning by developing models for a C++ based simulation platform called GridLAB-D. The examination of the feasibility, accuracy and stability of such systems without neglecting or jeopardizing security issues was also under serious consideration. Moreover, the restructuring of the power market and the upcoming "smartness" of the electricity infrastructure has brought up new challenges regarding security and stability issues as well as the handling of the vast amount of data produced in real time. Consequently, the use Big Data analytics along with the use of Artificial Intelligence techniques such as Neural Networks (NNs) and Fuzzy Logic (FL) in order to accurately anticipate, prevent or fix risky conditions was of high priority.

Degree

thesis:*
Grantor dc:publisher
University of Thessaly (UTH)
Year dc:date
2022

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Fainti, Rafik
  • Φάιντι, Ραφίκ

Subjects

dc:subject × 28

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/52969

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

Fainti, Rafik; Φάιντι, Ραφίκ. Platforms for (self) evolving smart energy systems. University of Thessaly (UTH), 2022. http://hdl.handle.net/10442/hedi/52969