Synthèse
Volume 30, Numéro 2, Pages 10-17
2025-12-01
Mathematical Models For Algerian Long Medium And Short Load Forecasting: Prevelec-dz And Prevos-dz
Authors : Khadir M. Tarek . Bougherira Ahmed .
Abstract
Electricity demand plays a critical role in the development of gro- wing economies. To meet daily electricity needs and support de- mand, utility companies and operators must utilize accurate and ef- ficient forecasting tools based on strong linear and nonlinear ma- thematical models. This paper will present briefly two tools deve- loped over the course of a 10 years collaboration period between the academic sector represented by the LabGED laboratory and the Algerian national electricity company (SONELGAZ) namely : PREVOS-DZ and PREVELC-DZ respectively for short/medium and long term forecasting. Short term load forecasting aid in both daily production planning where long-term forecasting helps deve- lopment strategies in terms of planning and construction of pro- duction means. The proposed software includes multiple regression methods as well as Artificial Neural Networks (ANNs) as basis for the large number of forecasting models in the software back end part. Forecasting capabilities extend beyond national demand to in- clude regional and special case requirements such as special days, Ramadhan etc. PREVOS-DZ and PREVELEC-DZ are Implemen- ted in C#, ensuring seamless compatibility with Excel spreadsheets, widely used by system operators. The predictive performance of fo- recasting models was validated using 2018 data, for short term fo- recasting. Part of the results are presented in this paper
Keywords
mathematical models ; artificial neural networks
Les articles similaires
Shaikh Ayub
.
More Satish
.
Bhirud Yuvaraj
.
Shelar Vaibhav
.
Bagwan Khalid
.
pages 7-32.
Sahed Abdelkader
.
Kahoui Hacene
.
pages 78-88.
Hawamdeh Younes
.
Yahiaoui Mohamed
.
pages 92-108.
Ahmed Mustafa Khudair
.
Abd Al Kareem Mustafa
.
pages 133-154.
