A bayesian-based neural network model for solar photovoltaic power forecasting

Angelo Ciaramella, Antonino Staiano, Guido Cervone, Stefano Alessandrini

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Solar photovoltaic power (PV) generation has increased constantly in several countries in the last ten years becoming an important component of a sustainable solution of the energy problem. In this paper, a methodology to 24 h or 48 h photovoltaic power forecasting based on a Neural Network, trained in a Bayesian framework, is proposed. More specifically, a multi-ahead prediction Multi-Layer Perceptron Neural Network is used, whose parameters are estimated by a probabilistic Bayesian learning technique. The Bayesian framework allows obtaining the confidence intervals and to estimate the error bars of the Neural Network predictions. In order to build an effective model for PV forecasting, the time series of Global Horizontal Irradiance, Cloud Cover, Direct Normal Irradiance, 2-m Temperature, azimuth angle and solar Elevation Angle are used and preprocessed by a Linear Predictive Coding technique. The experimental results show a low percentage of forecasting error on test data, which is encouraging if compared to state-of-the-art methods in literature.

Original languageEnglish (US)
Title of host publicationAdvances in Neural Networks - Computational Intelligence for ICT
EditorsAnna Esposito, Anna Esposito, Francesco Carlo Morabito, Eros Pasero, Simone Bassis
PublisherSpringer Science and Business Media Deutschland GmbH
Pages169-177
Number of pages9
ISBN (Print)9783319337463
DOIs
StatePublished - Jan 1 2016
EventInternational Workshop on Neural Networks, WIRN 2015 - Vietri sul Mare, Italy
Duration: May 20 2015May 22 2015

Publication series

NameSmart Innovation, Systems and Technologies
Volume54
ISSN (Print)2190-3018
ISSN (Electronic)2190-3026

Other

OtherInternational Workshop on Neural Networks, WIRN 2015
CountryItaly
CityVietri sul Mare
Period5/20/155/22/15

Fingerprint

Neural networks
Multilayer neural networks
Power generation
Time series
Network model
Prediction
Temperature
Bayesian learning
Methodology
Forecasting error
Confidence interval
Energy
Build-to-order

All Science Journal Classification (ASJC) codes

  • Decision Sciences(all)
  • Computer Science(all)

Cite this

Ciaramella, A., Staiano, A., Cervone, G., & Alessandrini, S. (2016). A bayesian-based neural network model for solar photovoltaic power forecasting. In A. Esposito, A. Esposito, F. C. Morabito, E. Pasero, & S. Bassis (Eds.), Advances in Neural Networks - Computational Intelligence for ICT (pp. 169-177). (Smart Innovation, Systems and Technologies; Vol. 54). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-319-33747-0_17
Ciaramella, Angelo ; Staiano, Antonino ; Cervone, Guido ; Alessandrini, Stefano. / A bayesian-based neural network model for solar photovoltaic power forecasting. Advances in Neural Networks - Computational Intelligence for ICT. editor / Anna Esposito ; Anna Esposito ; Francesco Carlo Morabito ; Eros Pasero ; Simone Bassis. Springer Science and Business Media Deutschland GmbH, 2016. pp. 169-177 (Smart Innovation, Systems and Technologies).
@inproceedings{a3e4e121e7cd4f8799fa0ab0d002e6d7,
title = "A bayesian-based neural network model for solar photovoltaic power forecasting",
abstract = "Solar photovoltaic power (PV) generation has increased constantly in several countries in the last ten years becoming an important component of a sustainable solution of the energy problem. In this paper, a methodology to 24 h or 48 h photovoltaic power forecasting based on a Neural Network, trained in a Bayesian framework, is proposed. More specifically, a multi-ahead prediction Multi-Layer Perceptron Neural Network is used, whose parameters are estimated by a probabilistic Bayesian learning technique. The Bayesian framework allows obtaining the confidence intervals and to estimate the error bars of the Neural Network predictions. In order to build an effective model for PV forecasting, the time series of Global Horizontal Irradiance, Cloud Cover, Direct Normal Irradiance, 2-m Temperature, azimuth angle and solar Elevation Angle are used and preprocessed by a Linear Predictive Coding technique. The experimental results show a low percentage of forecasting error on test data, which is encouraging if compared to state-of-the-art methods in literature.",
author = "Angelo Ciaramella and Antonino Staiano and Guido Cervone and Stefano Alessandrini",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-3-319-33747-0_17",
language = "English (US)",
isbn = "9783319337463",
series = "Smart Innovation, Systems and Technologies",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "169--177",
editor = "Anna Esposito and Anna Esposito and Morabito, {Francesco Carlo} and Eros Pasero and Simone Bassis",
booktitle = "Advances in Neural Networks - Computational Intelligence for ICT",
address = "Germany",

}

Ciaramella, A, Staiano, A, Cervone, G & Alessandrini, S 2016, A bayesian-based neural network model for solar photovoltaic power forecasting. in A Esposito, A Esposito, FC Morabito, E Pasero & S Bassis (eds), Advances in Neural Networks - Computational Intelligence for ICT. Smart Innovation, Systems and Technologies, vol. 54, Springer Science and Business Media Deutschland GmbH, pp. 169-177, International Workshop on Neural Networks, WIRN 2015, Vietri sul Mare, Italy, 5/20/15. https://doi.org/10.1007/978-3-319-33747-0_17

A bayesian-based neural network model for solar photovoltaic power forecasting. / Ciaramella, Angelo; Staiano, Antonino; Cervone, Guido; Alessandrini, Stefano.

Advances in Neural Networks - Computational Intelligence for ICT. ed. / Anna Esposito; Anna Esposito; Francesco Carlo Morabito; Eros Pasero; Simone Bassis. Springer Science and Business Media Deutschland GmbH, 2016. p. 169-177 (Smart Innovation, Systems and Technologies; Vol. 54).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A bayesian-based neural network model for solar photovoltaic power forecasting

AU - Ciaramella, Angelo

AU - Staiano, Antonino

AU - Cervone, Guido

AU - Alessandrini, Stefano

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Solar photovoltaic power (PV) generation has increased constantly in several countries in the last ten years becoming an important component of a sustainable solution of the energy problem. In this paper, a methodology to 24 h or 48 h photovoltaic power forecasting based on a Neural Network, trained in a Bayesian framework, is proposed. More specifically, a multi-ahead prediction Multi-Layer Perceptron Neural Network is used, whose parameters are estimated by a probabilistic Bayesian learning technique. The Bayesian framework allows obtaining the confidence intervals and to estimate the error bars of the Neural Network predictions. In order to build an effective model for PV forecasting, the time series of Global Horizontal Irradiance, Cloud Cover, Direct Normal Irradiance, 2-m Temperature, azimuth angle and solar Elevation Angle are used and preprocessed by a Linear Predictive Coding technique. The experimental results show a low percentage of forecasting error on test data, which is encouraging if compared to state-of-the-art methods in literature.

AB - Solar photovoltaic power (PV) generation has increased constantly in several countries in the last ten years becoming an important component of a sustainable solution of the energy problem. In this paper, a methodology to 24 h or 48 h photovoltaic power forecasting based on a Neural Network, trained in a Bayesian framework, is proposed. More specifically, a multi-ahead prediction Multi-Layer Perceptron Neural Network is used, whose parameters are estimated by a probabilistic Bayesian learning technique. The Bayesian framework allows obtaining the confidence intervals and to estimate the error bars of the Neural Network predictions. In order to build an effective model for PV forecasting, the time series of Global Horizontal Irradiance, Cloud Cover, Direct Normal Irradiance, 2-m Temperature, azimuth angle and solar Elevation Angle are used and preprocessed by a Linear Predictive Coding technique. The experimental results show a low percentage of forecasting error on test data, which is encouraging if compared to state-of-the-art methods in literature.

UR - http://www.scopus.com/inward/record.url?scp=84977138112&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84977138112&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-33747-0_17

DO - 10.1007/978-3-319-33747-0_17

M3 - Conference contribution

AN - SCOPUS:84977138112

SN - 9783319337463

T3 - Smart Innovation, Systems and Technologies

SP - 169

EP - 177

BT - Advances in Neural Networks - Computational Intelligence for ICT

A2 - Esposito, Anna

A2 - Esposito, Anna

A2 - Morabito, Francesco Carlo

A2 - Pasero, Eros

A2 - Bassis, Simone

PB - Springer Science and Business Media Deutschland GmbH

ER -

Ciaramella A, Staiano A, Cervone G, Alessandrini S. A bayesian-based neural network model for solar photovoltaic power forecasting. In Esposito A, Esposito A, Morabito FC, Pasero E, Bassis S, editors, Advances in Neural Networks - Computational Intelligence for ICT. Springer Science and Business Media Deutschland GmbH. 2016. p. 169-177. (Smart Innovation, Systems and Technologies). https://doi.org/10.1007/978-3-319-33747-0_17