LAPSE:2018.0511
Published Article
LAPSE:2018.0511
Deep Learning to Forecast Solar Irradiance Using a Six-Month UTSA SkyImager Dataset
September 21, 2018
Distributed PV power generation necessitates both intra-hour and day-ahead forecasting of solar irradiance. The UTSA SkyImager is an inexpensive all-sky imaging system built using a Raspberry Pi computer with camera. Reconfigurable for different operational environments, it has been deployed at the National Renewable Energy Laboratory (NREL), Joint Base San Antonio, and two locations in the Canary Islands. The original design used optical flow to extrapolate cloud positions, followed by ray-tracing to predict shadow locations on solar panels. The latter problem is mathematically ill-posed. This paper details an alternative strategy that uses artificial intelligence (AI) to forecast irradiance directly from an extracted subimage surrounding the sun. Several different AI models are compared including Deep Learning and Gradient Boosted Trees. Results and error metrics are presented for a total of 147 days of NREL data collected during the period from October 2015 to May 2016.
Record ID
Keywords
all-sky imaging, Artificial Intelligence, decision tree learning, deep learning, optical flow, solar irradiance forecasting
Subject
Suggested Citation
Moncada A, Richardson W Jr, Vega-Avila R. Deep Learning to Forecast Solar Irradiance Using a Six-Month UTSA SkyImager Dataset. (2018). LAPSE:2018.0511
Author Affiliations
Moncada A: Department of Mathematics, University of Texas at San Antonio, San Antonio, TX 78249, USA
Richardson W Jr: Department of Mathematics, University of Texas at San Antonio, San Antonio, TX 78249, USA [ORCID]
Vega-Avila R: CPS Energy, San Antonio, TX 78205, USA
[Login] to see author email addresses.
Richardson W Jr: Department of Mathematics, University of Texas at San Antonio, San Antonio, TX 78249, USA [ORCID]
Vega-Avila R: CPS Energy, San Antonio, TX 78205, USA
[Login] to see author email addresses.
Journal Name
Energies
Volume
11
Issue
8
Article Number
E1988
Year
2018
Publication Date
2018-07-31
ISSN
1996-1073
Version Comments
Original Submission
Other Meta
PII: en11081988, Publication Type: Journal Article
Record Map
Published Article
LAPSE:2018.0511
This Record
External Link
https://doi.org/10.3390/en11081988
Publisher Version
Download
Meta
Record Statistics
Record Views
845
Version History
[v1] (Original Submission)
Sep 21, 2018
Verified by curator on
Sep 21, 2018
This Version Number
v1
Citations
Most Recent
This Version
URL Here
https://psecommunity.org/LAPSE:2018.0511
Record Owner
Calvin Tsay
Links to Related Works