Hybrid mechanism-data-driven dynamic modeling for battery energy storage systems in primary frequency regulation
Year: 2026
Type: article
Abstract: The high penetration of renewable energy sources poses significant challenges to power system frequency stability. The battery energy storage systems (BESSs) with fast response characteristics, are playing an increasingly crucial role in primary frequency regulation (PFR). This paper proposes a hybrid mechanism-data-driven approach for the dynamic modeling of BESSs in PFR, aiming to enhance model accuracy under parameter varying scenarios. First, a hybrid modeling framework is established by classifying parameters into the inherent static and dynamic groups according to their time-varying characteristics. Specifically, the dynamic parameter representing the frequency regulation capability of BESSs varies rapidly, while the inherent static parameters evolve more gradually. Then, these inherent static parameters are identified using an auto-regressive moving average with exogenous inputs model combined with the forgetting factor recursive least squares method. Moreover, the state of charge is estimated through a multi-head attention-driven Transformer, which in turn enables the determination of the actual frequency regulation capability. Finally, case study results demonstrate that the root mean square error in frequency dynamic response of the obtained BESS model can be reduced by 85.4%, validating the superiority of the proposed modeling approach. © 2017 Elsevier Inc. All rights reserved.
Source: Journal of Energy Storage
Authors: Sen Yu, Yongji Cao, Changgang Li, Jiawen Cao, Meng Liu
Language: English
Subfield: Energy Engineering and Power Technology
Field: Energy
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,560
Awards: G1855082012, G3854423799
OpenAlex ID: W7169844718