Back to search

Aristotle University Of Thessaloniki (AUTH)

Διαμόρφωση τυπικών καμπυλών ηλεκτρικού φορτίου και εφαρμογή σε θέματα πρόβλεψης φορτίου και απόκρισης της ζήτησης

Abstract

dc:description

Due to the continually changing scene in the planning and operation of electrical energy systems there is a need for the efficient treatment of many issues, for example the large increase of the demand, the optimal management of generation capacity, the effect of the power sector in the environment and others. Considerable part of the electrical energy systems related research recognizes the important role of the consumer in the new arising scene. Under this approach, the research community seeks for tools for the characterization of the energy patterns of the various consumer types. The clustering of the patterns is a first step for their characterization. The main scope of the present doctoral dissertation is the development of a model for the clustering of the energy pattern of various consumers. The clustering is formulated as an exclusively unsupervised machine learning problem. The patterns are distinguished based on the level of similarity they present, where external information about the optimal number of clusters is absent. A comparative analysis of all clustering algorithms that have been proposed in the literature takes place. High efficient algorithms that already used in other problems are proposed and new algorithms that improve the operation of existing ones are developed. The comparative analysis considers all the validity indicators of the algorithms. New indicators are examined, alternative forms of the existing indicators are recommended and new criteria for the algorithm comparisons are also recommended. Moreover, all the representation techniques of the literature via the load curves are expressed and examined and a new technique is recommended. Finally, methods for the reduction of the amount of the data are recommended. Next, the problem of the short-term bus load forecasting is treated through the context of the load curves clustering. The forecast refers to the extraction of the next day load curve or the next hour load. Models based on artificial neural networks are developed and hybrid models that combine neural networks and the clustering of loads. Additionally, the application of clustering in the formulation of dynamic tariffs is investigated. Dynamic pricing is a fundamental pillar of demand response. In the present dissertation the application of demand response measures is done through a retailer, where the objective is to maximize the profitability in a deregulated market environment. The retailer`s actions in the market are simulated via a model which simulate the demand response of the consumers to the dynamic tariffs, identifies the optimal procurement mechanism for serving the load and examines the influence of various factors in the profitability.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2015

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Panapakidis, Ioannis
  • Παναπακίδης, Ιωάννης

Subjects

dc:subject × 12

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/37874

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

Panapakidis, Ioannis; Παναπακίδης, Ιωάννης. Διαμόρφωση τυπικών καμπυλών ηλεκτρικού φορτίου και εφαρμογή σε θέματα πρόβλεψης φορτίου και απόκρισης της ζήτησης. Aristotle University Of Thessaloniki (AUTH), 2015. http://hdl.handle.net/10442/hedi/37874