Primary Frequency Modulation Control of Advanced Adiabatic Compressed Air Energy Storage Based on Modified Dynamic Mode Decomposition

LI Tianyu; CHEN Laijun; CUI Sen; LIU Hanchen; MEI Shengwei

doi:10.13334/j.0258-8013.pcsee.250633

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Primary Frequency Modulation Control of Advanced Adiabatic Compressed Air Energy Storage Based on Modified Dynamic Mode Decomposition

更新时间：2026-09-28

- Primary Frequency Modulation Control of Advanced Adiabatic Compressed Air Energy Storage Based on Modified Dynamic Mode Decomposition
- Vol. 46, Issue 14, Pages: 5869-5878(2026)
- 作者机构：
  
  清华大学电机工程与应用电子技术系，北京市海淀区,100084
- 作者简介：
- 基金信息：
- DOI：10.13334/j.0258-8013.pcsee.250633
  CLC：
- Published：2026
- 稿件说明：
移动端阅览
LI Tianyu, CHEN Laijun, CUI Sen, et al. Primary Frequency Modulation Control of Advanced Adiabatic Compressed Air Energy Storage Based on Modified Dynamic Mode Decomposition[J]. 2026, 46(14): 5869-5878.
DOI：

LI Tianyu, CHEN Laijun, CUI Sen, et al. Primary Frequency Modulation Control of Advanced Adiabatic Compressed Air Energy Storage Based on Modified Dynamic Mode Decomposition[J]. 2026, 46(14): 5869-5878. DOI： 10.13334/j.0258-8013.pcsee.250633.

摘要

先进绝热压缩空气储能(advanced adiabatic compressed air energy storage，AA-CAES)作为一种大容量惯性系统，能够有效提升电网的频率调节能力。然而，AA-CAES的非线性运行特征限制了系统反馈控制律的整定。该文针对AA- CAES的频率调节问题，提出一种数据驱动的线性二次型调节器(data-driven linear regulation controller，DLQR)。首先，基于Koopman算符(Koopman operator，KO)理论提出改进动态模态分解算法(modified dynamic mode decomposition algorithm，MDMD)，全局线性化AA-CAES动态模型；其次，结合AA-CAES动态物理信息甄别MDMD观测函数以提升算法学习性能；随后，基于MDMD提出面向AA-CAES一次调频控制的DLQR，优化系统反馈控制律；最后，通过MATLAB/SIMULINK搭建AA-CAES动态模型进行算例分析。结果表明，基于MDMD的DLQR能够准确描述AA- CAES的动态行为并给出最优控制律，且在频率调节过程中所提方法能够有效跟踪功率指令并显著改善AA-CAES的频率响应特性。

Abstract

Advanced adiabatic compressed air energy storage (AA-CAES)

as a large-capacity system

can effectively boost the frequency regulation capacity of power grids. Nevertheless

the nonlinear features of AA-CAES constrain the rectification of the feedback control law. This paper presents a data-driven linear quadratic regulator (DLQR) for AA-CAES to provide temporal frequency support. Firstly

based on the Koopman operator (KO) theory

a modified dynamic mode decomposition (MDMD) algorithm is put forward

which can globally linearize the AA-CAES model. Secondly

by incorporating the physical knowledge of AA-CAES into the selection of the observables

the performance of dynamic learning is enhanced. Then

a DLQR based on MDMD is designed to optimize the feedback control law of the system. Finally

the AA-CAES dynamic model is constructed in MATLAB/SIMULINK for simulation and analysis. Case studies demonstrate that the DLQR can accurately describe the dynamic behavior of AA-CAES and provide the optimal control law. During the frequency regulation process

the proposed method can effectively track the power command and significantly improve the frequency response characteristics of AA-CAES.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Real-time Optimal Scheduling of Cascaded Hydropower Stations Under High-intensity Peak Shaving and Frequency Regulation

Modulation Wave and Harmonic Analysis of Single Active Vector PWM at Low Carrier Ratio

Product Mode Regulation and Application of Surface Dielectric Barrier Discharge Based on Electrode Temperature Control

Multi-condition Dynamic Degradation Characteristics of PEMFC Using TCN-LSTM and Transfer Learning

Topology Adaptability Evaluation of Hybrid HVDC Converter Stations Based on Reliability and Economic Analysis

Related Author

No data

Related Institution

No data

AI问答

⁰