TY - GEN
T1 - Comparing Spatio-Temporal Models for Aggregate PV Power Nowcasting
AU - Ruan, Guoping
AU - Chen, Xiaoyang
AU - Du, Yang
AU - Lim, Eng Gee
AU - Fang, Lurui
AU - Yan, Ke
N1 - Funding Information:
ACKNOWLEDGMENT This work is partially supported by the XJTLU Research Development Funding RDF-21-02-016
Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - The photovoltaic (PV) power fluctuations caused by passing clouds have become a major concern for grid operators. Consequently, utilities are requiring proper treatments to limit the intermittent PV generation. On this point, solar nowcasting provides a remedy by enabling the transition from reactive control to proactive, which often offers remarkable reliability to PV systems. Sensor networks that utilize spatio-temporal models are considered promising for solar nowcasting. However, current studies on sensor network nowcasting have dedicated much to point nowcasts, where the geographic smoothing effect that occurs in aggregate PV systems is generally left out. In this context, this paper presents a comparing study on spatio-temporal models for aggregate PV power nowcasting. Through empirical studies, the forecast skill of spatio-temporal models is found to decrease for a larger PV aggregation. In addition, the spatio-temporal regression shows competitive performance in various scenarios, yielding a priority for practical use.
AB - The photovoltaic (PV) power fluctuations caused by passing clouds have become a major concern for grid operators. Consequently, utilities are requiring proper treatments to limit the intermittent PV generation. On this point, solar nowcasting provides a remedy by enabling the transition from reactive control to proactive, which often offers remarkable reliability to PV systems. Sensor networks that utilize spatio-temporal models are considered promising for solar nowcasting. However, current studies on sensor network nowcasting have dedicated much to point nowcasts, where the geographic smoothing effect that occurs in aggregate PV systems is generally left out. In this context, this paper presents a comparing study on spatio-temporal models for aggregate PV power nowcasting. Through empirical studies, the forecast skill of spatio-temporal models is found to decrease for a larger PV aggregation. In addition, the spatio-temporal regression shows competitive performance in various scenarios, yielding a priority for practical use.
KW - Grid integration
KW - Photovoltaic power
KW - Solar forecasting
KW - Spatio-temporal models
UR - http://www.scopus.com/inward/record.url?scp=85146893585&partnerID=8YFLogxK
U2 - 10.1109/ISGTAsia54193.2022.10003491
DO - 10.1109/ISGTAsia54193.2022.10003491
M3 - Conference Proceeding
AN - SCOPUS:85146893585
T3 - Proceedings of the 11th International Conference on Innovative Smart Grid Technologies - Asia, ISGT-Asia 2022
SP - 580
EP - 584
BT - Proceedings of the 11th International Conference on Innovative Smart Grid Technologies - Asia, ISGT-Asia 2022
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 11th International Conference on Innovative Smart Grid Technologies - Asia, ISGT-Asia 2022
Y2 - 1 November 2022 through 5 November 2022
ER -