Hybrid approaches based on deep whole-sky-image learning to photovoltaic generation forecasting

Kong，Weicong1,2; Jia，Youwei1; Dong，Zhao Yang2; Meng，Ke2; Chai，Songjian3

2020-12-15

10.1016/j.apenergy.2020.115875

APPLIED ENERGY   影响因子和分区

0306-2619

1872-9118

With the ever-increased penetration of solar energy in the power grid, solar photovoltaic forecasting has become an indispensable aspect in maintaining power system stability and economic operation. At the operating stage, the forecasting accuracy of renewables has a direct influence on energy scheduling and dispatching. In this paper, we propose a series of novel approaches based on deep whole-sky-image learning architectures for very short-term solar photovoltaic generation forecasting, of which the lookahead windows concern the scales from 4 to 20 min. In particular, multiple deep learning models with the integration of both static sky image units and dynamic sky image stream are explicitly investigated. Extensive numerical studies on various models are carried out, through which the experimental results show that the proposed hybrid static image forecaster provides superior performance as compared to the benchmarking methods (i.e. the ones without sky images), with up to 8.3% improvement in general, and up to 32.8% improvement in the cases of ramp events. In addition, case studies at multiple time scales reveal that sky-image-based models can be more robust to the ramp events in solar photovoltaic generation.

Convolutional LSTM Deep learning Solar generation forecasting Whole Sky image

SCI ; EI

英语

第一 ; 通讯

Australian Research Council (ARC)["DP170103427","DP180103217"] ; Natural Science Foundation of China[72071100] ; Guangdong Basic and Applied Basic Research Fund[2019A1515111173] ; Dept. of Education of Guangdong Province, Young Talent Program[2018KQNCX223] ; High-level University Fund[G02236002]

Energy & Fuels ; Engineering

Energy & Fuels ; Engineering, Chemical

WOS:000594131000001

ELSEVIER SCI LTD

20204509453551

Image enhancement ; Long short-term memory ; Benchmarking ; Convolution ; Solar power generation ; System stability ; Electric power transmission networks ; Numerical methods ; Solar energy

Solar Power:615.2 ; Solar Energy and Phenomena:657.1 ; Electric Power Transmission:706.1.1 ; Information Theory and Signal Processing:716.1 ; Numerical Methods:921.6 ; Systems Science:961

ENGINEERING

2-s2.0-85094919413

Scopus

1.Department of Electrical and Electronic Engineering,Southern University of Science and Technology,China
2.School of Electrical Engineering and Telecommunications,the University of New South Wales,Sydney,Australia
3.Department of Electrical Engineering,The Hong Kong Polytechnic University,Hong Kong

Kong，Weicong,Jia，Youwei,Dong，Zhao Yang,et al. Hybrid approaches based on deep whole-sky-image learning to photovoltaic generation forecasting[J]. APPLIED ENERGY,2020,280.

Kong，Weicong,Jia，Youwei,Dong，Zhao Yang,Meng，Ke,&Chai，Songjian.(2020).Hybrid approaches based on deep whole-sky-image learning to photovoltaic generation forecasting.APPLIED ENERGY,280.

Kong，Weicong,et al."Hybrid approaches based on deep whole-sky-image learning to photovoltaic generation forecasting".APPLIED ENERGY 280(2020).