@article{Xie2025, 
author = {Yan Xie and Shiying Ma and Jiakai Shen and Xiaojun Tang and Jianmiao Ren and Liwen Zheng},
title = {Wide-area Active Frequency Control with Multi-Step-Size MPC},
year = {2025},
journal = {CSEE Journal of Power and Energy Systems},
volume = {11},
number = {3},
pages = {972-986},
keywords = {Active control, frequency stability, model predictive control, multi-step-size, spatial temporal distribution},
url = {https://www.sciopen.com/article/10.17775/CSEEJPES.2024.06280},
doi = {10.17775/CSEEJPES.2024.06280},
abstract = {The construction of UHV AC/DC hybrid power grids and the integration of large-scale renewable energy have led to significant frequency stability issues. To enhance the frequency regulation capacity within areas and fully utilize the mutual supportive capacity among areas, active frequency control (AFC) theory has been proposed and developed based on the concept of active feed-forward control. However, existing AFC methods have limitations in computational solutions and control effects in large-scale, wide-area interconnected power systems. To address these shortcomings, this paper proposes an improved AFC method based on multi-step-size model predictive control (MSS-MPC). Through multi-step-size discretization, an improved model predictive control algorithm for AFC is derived based on the iterative solution of the multi-step-size linear quadratic regulator model, which ensures control accuracy through precise prediction and enhances the control performance through additional prediction. Based on the collaboration of two-level dispatching agencies and the consideration of differences in control objectives among areas at different stages, an improved three-stage AFC strategy is proposed, which aims to suppress additional frequency disturbances after control model switching by proposing the active-passive switching strategy. Case studies demonstrate that, compared with the existing AFC method, the proposed AFC method with the MSS-MPC algorithm has a better control effect and better computational performance.}
}