Depot Institutionnel de l'UMBB >
Thèses de Doctorat et Mémoires de Magister >
Génie Eléctriques >
Doctorat >
Veuillez utiliser cette adresse pour citer ce document :
http://dlibrary.univ-boumerdes.dz:8080/handle/123456789/7502
|
Titre: | Application of optimization to data communication in smart grids |
Auteur(s): | Saoud, Afaf Recioui, Abdelmadjid(Directeur de thèse) |
Mots-clés: | Smart grids Load forecasting Fog-cloud computing |
Date de publication: | 2021 |
Editeur: | Université M'Hamed Bougara : Institut de génie électrique et électronique |
Résumé: | Smart grid has been introduced as a new generation of power systems that ensures
reliable, secure, low cost, and intelligent energy distribution and consumption. In smart
grids, a complex two-way communication infrastructure is involved generating huge
amount of data from the different parts of the grid which generates delay and accuracy
problems that affect the performance of the smart grid.
In this thesis, optimization is applied to data communication at different levels of the
smart grid. Three significant issues are investigated: data transfer improvement in wide
area monitoring (WAMS), load balancing in cloud-fog computing and load energy
forecasting based on smart metering system data.
First, we propose an optimization of the WAMS data transfer through PMU reporting
rate. The objective of this work is based on the variation of the reporting rate to prove its
relation with the PMU location and compare the results to those of the fixed reporting rate
as specified in the standards. The Search Group Algorithm is used for the reporting rate
optimization. We consider the PMU data latency and completeness as performance
metrics. The simulation is performed on MATLAB/SIMULINK.
Second, load balancing in smart grid to overcome the delay issue is proposed. In this
work, we introduce a cloud-fog computing system and hybrid optimization based on
WOA-BAT to enhance the task scheduling in the virtual machines. The performance
measures for this study are the processing and response times. The simulation is carried out
on Java platform in Net beans and cloud analyst tool.
Finally, optimization applied to short term forecasting as an application on smart
metering data is presented. In this part, we optimize a long short-term memory autoencoder
(LSTM-AE) model parameters using Particle Swarm Optimization (PSO) to give
better results in terms of forecasting and then compare to state of art forecasting models.
The evaluation metrics used for the comparison are mean absolute error (MAE) and root
mean square error (RMSE). The simulation is done on PYTHON |
Description: | 131 p. : ill. ; 30 cm |
URI/URL: | http://dlibrary.univ-boumerdes.dz:8080/handle/123456789/7502 |
Collection(s) : | Doctorat Doctorat
|
Fichier(s) constituant ce document :
|
Tous les documents dans DSpace sont protégés par copyright, avec tous droits réservés.
|