Prediction of Daily Photovoltaic Energy Production Using Weather Data and Regression

Hüseyin Sarper, Igor Melnykov, Lee Anne Martínez

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

This paper presents linear regression models to predict the daily energy production of three photovoltaic (PV) systems located in southeast Virginia. The prediction is based on daylight duration, sky index, average relative humidity, and the presence of fog or mist. No other daily weather report components were statistically significant. The proposed method is easy to implement, and it can be used in conjunction with other advanced methods in estimating any given future day’s energy production if weather prediction is available. Data from 2013 to 2015 were used in the model construction. Model validation was performed using newer (2016, 2017, 2020, and 2021) data not used in the model construction. Results show good prediction accuracy for a simple methodology, free of system parameters, that can be utilized by ordinary photovoltaic energy users. The majority of the data was collected at the Old Dominion University. The entire data set can be downloaded using the link provided.

Original languageEnglish (US)
Article number064502
JournalJournal of Solar Energy Engineering, Transactions of the ASME
Volume143
Issue number6
DOIs
StatePublished - Dec 2021
Externally publishedYes

Bibliographical note

Publisher Copyright:
Copyright © 2021 by ASME.

Keywords

  • Daily solar energy
  • Energy prediction
  • Photovoltaics
  • Regression
  • Renewable energy
  • Weather data

Fingerprint

Dive into the research topics of 'Prediction of Daily Photovoltaic Energy Production Using Weather Data and Regression'. Together they form a unique fingerprint.

Cite this