Graph Convolutional Network-Based Topology Embedded Deep Reinforcement Learning for Voltage Stability Control

doi:10.1109/TPWRS.2021.3084469

Graph Convolutional Network-Based Topology Embedded Deep Reinforcement Learning for Voltage Stability Control

2021-09-01

66

OA

AI

DOI：10.1109/TPWRS.2021.3084469

原文链接

查看原文

摘要

En 中文

Topology changes happen frequently in power systems and can impose significant challenges to traditional controllers of power systems. Recent studies revealed the strength of deep reinforcement learning (DRL) based approaches in power system preventive and corrective control. But topological variations are difficult to capture using classical fully connected neural network (FCN) model and has not been explicitly modeled in previous work. Hence, we develop a Graph Convolutional Network (GCN) based DRL framework to tackle topology changes in power system voltage stability control design. The GCN model helps the DRL agent to better capture topology changes and spatial correlations in nodal features. Our GCN based approach is evaluated using the IEEE-39 bus system and it outperforms the FCN-based DRL scheme in terms of training convergence and control performance considering grid topology changes.

Keyword:

Topology

Network topology

Power system stability

Voltage control

Training

Feature extraction

Load shedding

Deep reinforcement learning

double-deep Q network

graph convolutional network

voltage control

Graph Convolutional Network-Based Topology Embedded Deep Reinforcement Learning for Voltage Stability Control

摘要

AI总结

期刊

机构

引用论文