What is it about?
This study improves short-term solar PV power forecasting by combining moving block bootstrap techniques with a feed-forward neural network. The model was tested using one year of real data from a 3.63 kWp photovoltaic system in Rio de Janeiro, Brazil, recorded every 30 minutes. Results showed that the proposed MBB-FFNN approach reduced forecasting errors compared with the original neural network model, supporting more reliable prediction of photovoltaic generation.
Featured Image
Photo by Andres Siimon on Unsplash
Why is it important?
This article can be useful for researchers, engineers, and students working on solar photovoltaic systems, renewable energy forecasting, and data-driven energy models. The proposed approach can be used as a reference for developing short-term PV power forecasting models using real operational data. It may also support studies focused on neural networks, bootstrap techniques, uncertainty reduction, and the integration of solar energy into power systems.
Perspectives
From my perspective, this publication contributes to the development of more reliable forecasting tools for photovoltaic energy systems. The use of real PV data was especially important because it allowed us to evaluate the proposed method under practical operating conditions, not only through theoretical assumptions. I consider this work relevant because accurate short-term forecasting can support better planning, reduce uncertainty, and strengthen the integration of solar energy into sustainable power systems.
Ph.D Eliseo Zarate-Perez
Universidad Privada del Norte
Read the Original
This page is a summary of: Optimization of Solar PV Power Forecasting Using Bootstrap Techniques and the Feed-Forward Neural Network Model, January 2022, LACCEI (Latin American and Caribbean Consortium of Engineering Institutions),
DOI: 10.18687/laccei2022.1.1.18.
You can read the full text:
Contributors
The following have contributed to this page
