Data-Driven Auxiliary Modeling and Efficient Solution of Renewable Power Based Unit Commitment Considering Dual Uncertainties
Energy Engineering · 2026 · 1-10
DOI: 10.32604/ee.2026.090474 ↗Kurz gefasst
Fachartikel in der Zeitschrift Energy Engineering, 2026. Verfasst von Congcong Liu und 6 weiteren Autor:innen.
Abstract (Originalsprache)
: The computational efficiency and accuracy of the unit commitment problem affect the safe and economic operation of power systems. With the continuous growth of renewable energy penetration, the volatility and uncertainty of such energy sources pose great challenges to the accurate formulation of power system unit commitment schedules. To address this issue, this paper proposes an efficient solution method for power system unit commitment considering dual uncertainties of power sources and loads. First, a nonparametric Dirichlet process Gaussian mixture model (DPGMM) combined with variational Bayesian inference is adopted to characterize the uncertainties of wind power, photovoltaic power and loads. Second, a data-driven solution model for unit commitment is established. In view of the time-period coupling characteristics of unit commitment, a feature vector construction method based on multi-head attention mechanism is utilized, and an improved Transformer neural network is proposed to pre-identify the on/off status of generating units. The physical unit commitment model is then employed for subsequent solution to guarantee feasibility. Finally, multiple standard IEEE test systems are used to verify the effectiveness and practicability of the proposed method. Simulation results on multiple IEEE test systems show that the proposed method achieves an average optimality gap below 1.25% and significantly shortens computation time; for the large-scale IEEE 2383-bus system, the solving time is reduced to 15.62 s, which demonstrates prominent advantages in both solution accuracy and computational efficiency.
Zitieren
APA
Liu, C., Li, P., Zhang, X., Chen, X., He, S., Zheng, C., & Yang, N. (2026). Data-Driven Auxiliary Modeling and Efficient Solution of Renewable Power Based Unit Commitment Considering Dual Uncertainties. Energy Engineering, 0(0), 1-10. https://doi.org/10.32604/ee.2026.090474
Vancouver
Liu C, Li P, Zhang X, Chen X, He S, Zheng C, et al. Data-Driven Auxiliary Modeling and Efficient Solution of Renewable Power Based Unit Commitment Considering Dual Uncertainties. Energy Engineering. 2026;0(0):1-10. doi:10.32604/ee.2026.090474
BibTeX
@article{Liu2026,
author = {Liu, Congcong and Li, Pengfei and Zhang, Xingyou and Chen, Xiaolu and He, Shuai and Zheng, Cai and Yang, Nan},
title = {Data-Driven Auxiliary Modeling and Efficient Solution of Renewable Power Based Unit Commitment Considering Dual Uncertainties},
journal = {Energy Engineering},
year = {2026},
volume = {0},
number = {0},
pages = {1--10},
doi = {10.32604/ee.2026.090474}
}
EUVIMED