arrow
返回

Soft Actor-Critic Reinforcement Learning for Reactive Current Injection Protocols

delete2026-01-01
delete0
PRE
AI
M
Mohana Fathollahi *
A
Antonio Camacho
C
Cecilio Ángulo
J
Jerrad Hampton
DOI:10.1007/978-3-032-11442-6_4delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
现代电力系统越来越受到对可再生能源依赖日益增长的挑战。由于这些能源固有的间歇性,它们可能导致电压波动和相不平衡,尤其是在电网扰动期间。在这些扰动中,电压骤降最为关键,其发生时间在秒级甚至毫秒级。一种必要的缓解策略涉及向电压降低的相注入无功电流。传统确定适当无功电流量的方法依赖电网规范,这些规范基于测量电压定义最小值;因此,这些方法缺乏优化,也无法适应以更好地满足电网需求。本研究提出了一种基于Soft Actor-Critic(SAC)的替代方案,这是一种无模型且离线的强化学习(RL)算法,旨在解决先前方法的不足。推理阶段的仿真结果表明,基于SAC的方法在性能上与基于优化的方法相当,同时能在毫秒级快速响应时间内更好地泛化至未见数据。
Keyword:
Reinforcement Learning
Soft Actor-Critic
Low-Voltage Ride-Through
Power Grid
Voltage Sag

期刊

A
ARTIFICIAL INTELLIGENCE XLII, AI 2025, PT II
IF:
0
论文数:
38
被引数:
0

机构

U
universitat politecnica de catalunya
学者数:
1.9W
论文数: 1.6W
被引数: 17
引用论文

引用论文

Improving Voltage Ride-Through Procedures in Distributed Generation Systems by Reinforcement Learning
err2024-09-25
err0
PREAI
errFathollahi,Mohana; Camacho,Antonio; Velasco,Manel; Martí,Pau; Angulo,Cecilio; Hampton,Jerrad
err分享
err收藏
Reinforcement Learning for Robust Voltage Control in Distribution Grid Under Uncertainties
err
err0
PREAI
errPetrusev,Aleksandr; Putratama,Muhammad Andy; Rigo-Mariani,Rémy; Debusschere,Vincent; Reignier,Patrick; Hadjsaid,Nouredine
err分享
err收藏
A Comprehensive Review of Control Strategies to Overcome Challenges During LVRT in PV Systems
err2021-01-01
err40
errOAAI
errJoshi, Jyoti; Swami, Anurag Kumar; Jately, Vibhu; Azzopardi, Brian
err分享
err收藏
Design Principles for Model Generalization and Scalable AI Integration in Radio Access Networks
err2025-01-01
err0
errOAAI
errSoldati, Pablo; Ghadimi, Euhanna; Demirel, Burak; Wang, Yu; Gaigalas, Raimundas; Sintorn, Mathias
err分享
err收藏
学者 查看更多内容