Normalized deep neural network with self-attention mechanism accelerated ADMM for distributed energy management of regional integrated energy systems considering renewable energy uncertainty

Linfei Yin, Rongkun Liu and Wei Ge

Energy, 2025, vol. 330, issue C

Abstract: Wind power and photovoltaic have strong volatility and strong uncertainty, and the energy management (EM) problems of the power system become increasingly complex. Traditional scheduling methods have a slow solution speed, which may be difficult to satisfy the real-time demands of EM. To address this problem, this study presents an alternating direction multiplier method (ADMM) for normalized deep neural networks based on self-attention mechanism (ADMM-NDNN-SAM). The proposed ADMM-NDNN-SAM is a mathematical-model dual-driven framework, where ADMM is a mathematical model and self-attention mechanism is a data-driven method. ADMM-NDNN-SAM organically combines the theoretical reliability of analytical methods with the flexibility and predictive capability of data-driven methods, and can perform intelligently and efficiently in dealing with large-scale and multi-constraint complex integrated energy system optimization problems. Experimental findings on IEEE 39- and 118-bus systems verify that:(1) computational speed is significantly improved. ADMM-NDNN-SAM accelerates the computation speed by 55.12 % and 52.90 % compared with ADMM; (2) compared to seven best-of-breed approaches, ADMM-NDNN-SAM reduces at least 7.11 % of the cost and 8.64 of the CO2 emissions on IEEE 39-bus system, and at least 0.03 % of the cost and 37.65 % of the CO2 emissions on IEEE 118-bus system; (3) scheduling scheme derived from the ADMM-NDNN-SAM method is robust.

Keywords: Distributed multi-objective energy management; Uncertainty; ADMM; Deep neural networks; Integrated energy systems (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544225025940
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:energy:v:330:y:2025:i:c:s0360544225025940

DOI: 10.1016/j.energy.2025.136952

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

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