Work

Forecasting Power System Flexibility Requirements: A Hybrid Deep-Learning Approach

Year: 2022

Type: article

Abstract: Variable and Renewable Energies (VRE) are growing fast among power systems all around the world. An important issue emerging along this fast development is to which extent a power system can cope with rapid changes either in load or generation. Thus, in future renewable-based power systems, power system flexibility requirements assessment and forecasting will become key tasks for grid operators, market participants as well as producers. In this paper, we investigate the use of a hybrid forecasting framework to forecast a given power system Residual Load (RL) function. Our forecasting framework relies on the decomposition of RL time-series into oscillatory patterns and the training of a Convolutional Long-Short Term Memory (ConvLSTM-2D) Artificial Neural Network (ANN). Using this methodology, we forecast every day of French region Occitanie power system hourly RL for the full year 2019, with an averaged MAPE score of 3.03%. This empirical application illustrates the consistent accuracy of forecasts given by our methodology over time, making it very suitable for daily practical use among energy stakeholders.

Source: Electric Power Systems Research

DOI: 10.1016/j.epsr.2024.111307

Author: Théotime Coudray

Institutions: Université de Montpellier, Acteurs, Ressources et Territoires dans le Développement

Language: English

Topic: Energy Load and Power Forecasting

Subfield: Electrical and Electronic Engineering

Field: Engineering

Domain: Physical Sciences

SDGs: Affordable and clean energy, Industry, innovation and infrastructure, Climate action

Open Access status: hybrid

Estimated APC paid: $2,580