Research Article Open Access

Analysis of Artificial Neural Network Point Forecasting Models and Prediction Intervals for Solar Irradiance Estimation

Antônio Fabrício Guimarães de Sousa1, Helaine Cristina Moraes Furtado1, Wilson Negrão Macêdo2 and Anderson Alvarenga de Moura Meneses1
  • 1 Federal University of Western Pará, Brazil
  • 2 Federal University of Pará, Brazil

Abstract

An accurate knowledge on solar irradiance prediction is particularly required for proper development and planning of Photovoltaic (PV) energy systems. The main purpose of the present research is to assess the accuracy of Artificial Neural Networks (ANN) short-term forecast of univariate solar irradiance time series, with conventional point prediction and Prediction Intervals (PIs), comparing models. The Lower Upper Bound Estimation trained with Particle Swarm Optimization (PSO-LUBE) was used for PIs estimation. Solar irradiance data collected from a station in Amazon region in Brazil was used to train and test the models. Results demonstrate that all ANN models yield good accuracy in terms of prediction error: 8.1-8.5% for normalized root Mean Square Error (nRMSE), 5.8-6.0% for normalized Mean Absolute Error (nMAE) and 94-95% for determination coefficient (R2). However, due to the accuracy of PI information (Coverage Probability = 94.94% and PI Normalized Average Width = 32.50%), PSO-LUBE was the best method tested for decision-making.

American Journal of Engineering and Applied Sciences
Volume 13 No. 3, 2020, 347-357

DOI: https://doi.org/10.3844/ajeassp.2020.347.357

Submitted On: 7 April 2020 Published On: 17 July 2020

How to Cite: de Sousa, A. F. G., Furtado, H. C. M., Macêdo, W. N. & Meneses, A. A. M. (2020). Analysis of Artificial Neural Network Point Forecasting Models and Prediction Intervals for Solar Irradiance Estimation. American Journal of Engineering and Applied Sciences, 13(3), 347-357. https://doi.org/10.3844/ajeassp.2020.347.357

  • 2,938 Views
  • 949 Downloads
  • 11 Citations

Download

Keywords

  • Solar Irradiance
  • Univariate Time Series Forecasting
  • Artificial Neural Networks
  • Prediction Intervals