Short-term reforecasting of power output from a 48 MWe solar PV plant

Yinghao Chu; Bryan Urquhart; Seyyed M.I. Gohari; Hugo T.C. Pedro; Jan Kleissl; Carlos F.M. Coimbra

doi:10.1016/j.solener.2014.11.017

Short-term reforecasting of power output from a 48 MWe solar PV plant

Yinghao Chu, Bryan Urquhart, Seyyed M.I. Gohari, Hugo T.C. Pedro, Jan Kleissl, Carlos F.M. Coimbra^*

^*Corresponding author for this work

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

237 Citations (Scopus)

Abstract

A smart, real-time reforecast method is applied to the intra-hour prediction of power generated by a 48 MWe photovoltaic (PV) plant. This reforecasting method is developed based on artificial neural network (ANN) optimization schemes and is employed to improve the performance of three baseline prediction models: (1) a physical deterministic model based on cloud tracking techniques; (2) an auto-regressive moving average (ARMA) model; and (3) a k-th Nearest Neighbor (kNN) model. Using the measured power data from the PV plant, the performance of all forecasts is assessed in terms of common error statistics (mean bias, mean absolute error and root mean square error) and forecast skill over the reference persistence model. With the reforecasting method, the forecast skills of the three baseline models are significantly increased for time horizons of 5, 10, and 15. min. This study demonstrates the effectiveness of the optimized reforecasting method in reducing learnable errors produced by a diverse set of forecast methodologies. © 2014 Elsevier Ltd.

Original language	English
Pages (from-to)	68-77
Journal	Solar Energy
Volume	112
DOIs	https://doi.org/10.1016/j.solener.2014.11.017
Publication status	Published - 1 Feb 2015
Externally published	Yes

Bibliographical note

Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

Research Keywords

Artificial neural networks
Genetic algorithm optimization
PV generation
Real-time reforecasting

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access Output

10.1016/j.solener.2014.11.017

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{b244b69d5f504428b6141342aa0e0638,

title = "Short-term reforecasting of power output from a 48 MWe solar PV plant",

abstract = "A smart, real-time reforecast method is applied to the intra-hour prediction of power generated by a 48 MWe photovoltaic (PV) plant. This reforecasting method is developed based on artificial neural network (ANN) optimization schemes and is employed to improve the performance of three baseline prediction models: (1) a physical deterministic model based on cloud tracking techniques; (2) an auto-regressive moving average (ARMA) model; and (3) a k-th Nearest Neighbor (kNN) model. Using the measured power data from the PV plant, the performance of all forecasts is assessed in terms of common error statistics (mean bias, mean absolute error and root mean square error) and forecast skill over the reference persistence model. With the reforecasting method, the forecast skills of the three baseline models are significantly increased for time horizons of 5, 10, and 15. min. This study demonstrates the effectiveness of the optimized reforecasting method in reducing learnable errors produced by a diverse set of forecast methodologies. {\textcopyright} 2014 Elsevier Ltd.",

keywords = "Artificial neural networks, Genetic algorithm optimization, PV generation, Real-time reforecasting",

author = "Yinghao Chu and Bryan Urquhart and Gohari, {Seyyed M.I.} and Pedro, {Hugo T.C.} and Jan Kleissl and Coimbra, {Carlos F.M.}",

note = "Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].",

year = "2015",

month = feb,

day = "1",

doi = "10.1016/j.solener.2014.11.017",

language = "English",

volume = "112",

pages = "68--77",

journal = "Solar Energy",

issn = "0038-092X",

publisher = "Elsevier Ltd.",

}

TY - JOUR

T1 - Short-term reforecasting of power output from a 48 MWe solar PV plant

AU - Chu, Yinghao

AU - Urquhart, Bryan

AU - Gohari, Seyyed M.I.

AU - Pedro, Hugo T.C.

AU - Kleissl, Jan

AU - Coimbra, Carlos F.M.

N1 - Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

PY - 2015/2/1

Y1 - 2015/2/1

N2 - A smart, real-time reforecast method is applied to the intra-hour prediction of power generated by a 48 MWe photovoltaic (PV) plant. This reforecasting method is developed based on artificial neural network (ANN) optimization schemes and is employed to improve the performance of three baseline prediction models: (1) a physical deterministic model based on cloud tracking techniques; (2) an auto-regressive moving average (ARMA) model; and (3) a k-th Nearest Neighbor (kNN) model. Using the measured power data from the PV plant, the performance of all forecasts is assessed in terms of common error statistics (mean bias, mean absolute error and root mean square error) and forecast skill over the reference persistence model. With the reforecasting method, the forecast skills of the three baseline models are significantly increased for time horizons of 5, 10, and 15. min. This study demonstrates the effectiveness of the optimized reforecasting method in reducing learnable errors produced by a diverse set of forecast methodologies. © 2014 Elsevier Ltd.

AB - A smart, real-time reforecast method is applied to the intra-hour prediction of power generated by a 48 MWe photovoltaic (PV) plant. This reforecasting method is developed based on artificial neural network (ANN) optimization schemes and is employed to improve the performance of three baseline prediction models: (1) a physical deterministic model based on cloud tracking techniques; (2) an auto-regressive moving average (ARMA) model; and (3) a k-th Nearest Neighbor (kNN) model. Using the measured power data from the PV plant, the performance of all forecasts is assessed in terms of common error statistics (mean bias, mean absolute error and root mean square error) and forecast skill over the reference persistence model. With the reforecasting method, the forecast skills of the three baseline models are significantly increased for time horizons of 5, 10, and 15. min. This study demonstrates the effectiveness of the optimized reforecasting method in reducing learnable errors produced by a diverse set of forecast methodologies. © 2014 Elsevier Ltd.

KW - Artificial neural networks

KW - Genetic algorithm optimization

KW - PV generation

KW - Real-time reforecasting

UR - http://www.scopus.com/inward/record.url?scp=84916618688&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-84916618688&origin=recordpage

U2 - 10.1016/j.solener.2014.11.017

DO - 10.1016/j.solener.2014.11.017

M3 - RGC 21 - Publication in refereed journal

SN - 0038-092X

VL - 112

SP - 68

EP - 77

JO - Solar Energy

JF - Solar Energy

ER -

Short-term reforecasting of power output from a 48 MWe solar PV plant

Abstract

Bibliographical note

Research Keywords

UN SDGs

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this