自动化学报2025，Vol.51Issue(10)：2337-2346,10.DOI:10.16383/j.aas.c250261
面向电力系统快速频率响应的数据与模型驱动预测控制
Data and Model-driven Predictive Control for Fast Frequency Response in Power Systems
摘要
Abstract
A key objective of power system control is to maintain frequency stability.However,high penetration of renewable energy resources may cause frequent power fluctuations,resulting in negative impact on system fre-quency regulation.To deal with this issue,it commonly requires to rapidly adjust the power output of inverter-based resources in response to system frequency fluctuations,achieving fast frequency control.This paper proposes a data and model-driven predictive control method for fast frequency control in power systems.First,a data-driven disturbance observer is designed to estimate system disturbance such as load changes,renewable energy fluctu-ations and so on.To optimize the control performance,the reference governor designed by a neural network provides virtual reference for the model predictive controller.The reference governor enhances the performance of the controller with a short prediction horizon through learning a long prediction horizon model predictive controller,thus reducing the required computation time.Finally,simulation comparison results demonstrate that the proposed method can effectively improve the frequency control performance.关键词
快速频率控制/变流器资源/模型预测控制/参考调节器/深度学习Key words
Fast frequency control/inverter-based resource/model predictive control/reference governor/deep learning引用本文复制引用
吴卓睿,张萌,管晓宏..面向电力系统快速频率响应的数据与模型驱动预测控制[J].自动化学报,2025,51(10):2337-2346,10.基金项目
国家自然科学基金(62033005,62273270),能源陕西实验室项目(ESLB202413)资助Supported by National Natural Science Foundation of China(62033005,62273270)and S&T Program of Energy Shaanxi Laboratory(ESLB202413) （62033005,62273270）
