NSDAR: A neural network-based model for similar day screening and electric load forecasting

Zongxi Jiang, Luliang Zhang and Tianyao Ji

Applied Energy, 2023, vol. 349, issue C, No S0306261923010115

Abstract: The load forecasting methods based on similar days has been widely studied in decades, where similar days were mostly used as an auxiliary means to improve model performance. The process of transitioning from similar days to load forecasting results requires the use of complex models, which are often opaque and lack interpretability. Thus, this paper proposes a new method called NSDAR (Neural-based Similar Days Auto Regression) based on the similar days analyses. Different from the conventional methods based on similarity of meteorological factors or simple policy for similar day screening, the NSDAR uses neural networks for similar day screening, which can detect many unconventional similar days that are difficult for people to notice, and in order to obtain interpretable load forecasting results, NSDAR uses a linear model to generate load forecasting results based on the similar day screening result. In addition, the model results can also be used for regional load analysis, load clustering, and other purposes. This paper guides the structural design of the model through theoretical derivation and verifies the feasibility of this method. NSDAR performs well in similar day screening and load forecasting tasks, and its results have better interpretability compared to the other existing load forecasting models.

Keywords: Short-term load forecasting; Similar days; Convolution network; Interpretable model (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0306261923010115
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:appene:v:349:y:2023:i:c:s0306261923010115

Ordering information: This journal article can be ordered from
http://www.elsevier.com/wps/find/journaldescription.cws_home/405891/bibliographic
http://www.elsevier. ... 405891/bibliographic

DOI: 10.1016/j.apenergy.2023.121647

Access Statistics for this article

Applied Energy is currently edited by J. Yan

More articles in Applied Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().