Load Forecasting Using Dynamic High-Order Neural Networks - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 1993

Load Forecasting Using Dynamic High-Order Neural Networks

Résumé

A new short-term load forecasting model based on recurrent high-order neural networks (RHONNs) is presented in this paper. Short-term load forecasts are required so that the dispatch and scheduling of the generating units connected to a power system can be performed to meet the expected demand as economically and securely as possible. The proposed load forecasting is able to provide accurate load forecasts by using the minimum of input data, that is only historic al load data. Due to its dynamic nature the RHONN forecasting model is able to adapt quickly to changing load conditions like important load variations of changes of the daily load pattern (weekdays, holidays etc). Simulation results over hourly load timeseries of the Greek island of Crete demonstrate the efficiency of the model. The aim of the paper is to illustrate the capacity of this new family of neural network to learn the dynamics of the time series considered. The focus is thus on the learning rather than the testing performance. Generalization over out of sample data will be assessed in forthcoming works.
Fichier principal
Vignette du fichier
1993_PowerTech_LoadForecast_Kariniotakis.pdf (3.2 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03821839 , version 1 (30-10-2023)

Identifiants

Citer

Georges Kariniotakis, Elias B. Kosmatopoulos, Manolis A. Christodoulou, Georges S. Stavrakakis. Load Forecasting Using Dynamic High-Order Neural Networks. IEEE Joint International Power Conference Athens Power Tech, Sep 1993, Athènes, Greece. pp.801-805, ⟨10.1109/APT.1993.673905⟩. ⟨hal-03821839⟩
16 Consultations
7 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More