The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks

Due to the increasing integration of new energy sources, the power system now exhibits low inertia, in which the broadband oscillation problem is increasingly significant in the face of the strong coupling of complex and variable power systems, and the current lack of uniform and effective mathemati...

Full description

Saved in:
Bibliographic Details
Main Authors: Jinggeng Gao, Yong Yang, Honglei Xu, Yingzhou Xie, Chen Zhou, Haiying Dong
Format: Article
Language:English
Published: MDPI AG 2024-12-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/17/23/6062
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846124238660960256
author Jinggeng Gao
Yong Yang
Honglei Xu
Yingzhou Xie
Chen Zhou
Haiying Dong
author_facet Jinggeng Gao
Yong Yang
Honglei Xu
Yingzhou Xie
Chen Zhou
Haiying Dong
author_sort Jinggeng Gao
collection DOAJ
description Due to the increasing integration of new energy sources, the power system now exhibits low inertia, in which the broadband oscillation problem is increasingly significant in the face of the strong coupling of complex and variable power systems, and the current lack of uniform and effective mathematical models and analysis methods. To solve this major problem, a broadband oscillation localization method based on the combination of compressed perception and graph attention network (GAT) is proposed. The method firstly uses the principle of compression perception to compress and transmit the oscillation time series data of the sub-station, reconstructs the compressed signal at the master station and aggregates the grid topology and node characteristic information to effectively reduce the redundancy of the oscillation data; reconstruction error is only 0.031, takes into account the balance of the samples and the effectiveness of the computation, and adopts the multi-attention mechanism and the cross-entropy loss function to improve the performance of the model training. Finally, the offline training and online evaluation model based on the GAT algorithm is constructed, and the accuracy of the model is up to 98.5%; and the results show that the method has a high positioning accuracy and a certain anti-noise ability at the same time.
format Article
id doaj-art-bad6a87078234eb49b70b9aa5ef4a287
institution Kabale University
issn 1996-1073
language English
publishDate 2024-12-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj-art-bad6a87078234eb49b70b9aa5ef4a2872024-12-13T16:25:56ZengMDPI AGEnergies1996-10732024-12-011723606210.3390/en17236062The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention NetworksJinggeng Gao0Yong Yang1Honglei Xu2Yingzhou Xie3Chen Zhou4Haiying Dong5State Grid Gansu Electric Power Research Institute, Lanzhou 730000, ChinaState Grid Gansu Electric Power Research Institute, Lanzhou 730000, ChinaState Grid Gansu Electric Power Company, Lanzhou 730000, ChinaState Grid Gansu Electric Power Research Institute, Lanzhou 730000, ChinaSchool of New Energy and Power Engineering, Lanzhou Jiaotong University, Lanzhou 730070, ChinaSchool of New Energy and Power Engineering, Lanzhou Jiaotong University, Lanzhou 730070, ChinaDue to the increasing integration of new energy sources, the power system now exhibits low inertia, in which the broadband oscillation problem is increasingly significant in the face of the strong coupling of complex and variable power systems, and the current lack of uniform and effective mathematical models and analysis methods. To solve this major problem, a broadband oscillation localization method based on the combination of compressed perception and graph attention network (GAT) is proposed. The method firstly uses the principle of compression perception to compress and transmit the oscillation time series data of the sub-station, reconstructs the compressed signal at the master station and aggregates the grid topology and node characteristic information to effectively reduce the redundancy of the oscillation data; reconstruction error is only 0.031, takes into account the balance of the samples and the effectiveness of the computation, and adopts the multi-attention mechanism and the cross-entropy loss function to improve the performance of the model training. Finally, the offline training and online evaluation model based on the GAT algorithm is constructed, and the accuracy of the model is up to 98.5%; and the results show that the method has a high positioning accuracy and a certain anti-noise ability at the same time.https://www.mdpi.com/1996-1073/17/23/6062power systemcompressed sensinggraphical attention networkswideband oscillation positioning
spellingShingle Jinggeng Gao
Yong Yang
Honglei Xu
Yingzhou Xie
Chen Zhou
Haiying Dong
The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
Energies
power system
compressed sensing
graphical attention networks
wideband oscillation positioning
title The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
title_full The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
title_fullStr The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
title_full_unstemmed The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
title_short The Wideband Oscillatory Localization Method Based on Combining Compressed Sensing and Graph Attention Networks
title_sort wideband oscillatory localization method based on combining compressed sensing and graph attention networks
topic power system
compressed sensing
graphical attention networks
wideband oscillation positioning
url https://www.mdpi.com/1996-1073/17/23/6062
work_keys_str_mv AT jinggenggao thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT yongyang thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT hongleixu thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT yingzhouxie thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT chenzhou thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT haiyingdong thewidebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT jinggenggao widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT yongyang widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT hongleixu widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT yingzhouxie widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT chenzhou widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks
AT haiyingdong widebandoscillatorylocalizationmethodbasedoncombiningcompressedsensingandgraphattentionnetworks