An Adaptive Photovoltaic Production Estimator Based on Artificial Neural Networks

Thumbnail Image

Files

CIRED 2019 - 1987.pdf (507.86 KB)

Paper number

1987

Working Group Number

Conference name

CIRED 2019

Conference date

3-6 June 2019

Conference location

Madrid, Spain

Peer-reviewed

Yes

Short title

Convener

Authors

Corsetti, Edoardo, RSE, Italy
Guagliardi, Antonio G., RSE, Italy
Sandroni, Carlo, RSE S.p.A, Italy

Abstract

The need to decarbonize the energy sector is producing a tremendous increasing of the introduction of renewable generators supplied by non-controllable energy sources. In particular, in those countries with a high sun radiation for a relevant period of the year there is an ongoing growth of photovoltaic generators (PV). In this paper we propose an algorithm to forecast the day-ahead photovoltaic production based on artificial neural networks which exploit as input the day-ahead weather forecast data like: solar irradiations, solar irradiation measured on the PV panel surface, temperature of the PV panel, atmospheric pressure, air humidity, air temperature, wind intensity and direction. The artificial neural networks are daily trained with a very limited set of measure data trends: at most 5 past days taken within the last 20 days, and it able to provide the forecast of a pv-field with 35 kWp installed in the RSE test facility of distribute generation.With respect to other day-ahead photovoltaic production forecast methods the proposed one it is particularly suited to be applied in small plants, typically microgrid, with limited capacities in terms of monitoring, computational ability and storing measures. The day-ahead photovoltaic production forecast algorithm is also able to adapt to the surrounding conditions as it is daily updated with the latest measures.

Table of content

Keywords

Publisher

AIM

Date

2019-06-03

Published in

CIRED 2019 Conference

Permanent link to this record

https://cired-repository.org/handle/20.500.12455/676
http://dx.doi.org/10.34890/903

ISSN

2032-9644

ISBN

978-2-9602415-0-1