Sci. Tech. Energ. Transition
Volume 78, 2023
|Number of page(s)||9|
|Published online||12 July 2023|
Prediction method of photovoltaic power based on combination of CEEMDAN-SSA-DBN and LSTM
College of Electrical and New Energy, China Three Gorges University, Yichang, Hubei 443002, China
* Corresponding author: firstname.lastname@example.org
Accepted: 19 June 2023
Aiming at the problem of high fluctuation and instability of photovoltaic power, a photovoltaic power prediction method combining two techniques has been proposed in this study. In this method, the fast correlation filtering algorithm has been used to extract the meteorological features having a strong correlation with photovoltaic power generation. The complete ensemble empirical mode decomposition with an adaptive noise model has been used to decompose the data into high and low-frequency components to reduce the data volatility. Then, the long short-term neural network and the deep confidence network were combined into a new prediction model to predict each component. Finally, the proposed combined photovoltaic power prediction method has been analyzed using an example and compared with the other prediction methods. The results show that the proposed combined prediction method has higher prediction accuracy.
Key words: Photovoltaic power prediction / Empirical mode decomposition / Deep confidence network / Fast correlation filtering algorithm
© The Author(s), published by EDP Sciences, 2023
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.