2022
DOI: 10.1016/j.ijepes.2021.107648
|View full text |Cite
|
Sign up to set email alerts
|

A method of multivariate short-term voltage stability assessment based on heterogeneous graph attention deep network

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
8
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1
1

Relationship

0
9

Authors

Journals

citations
Cited by 17 publications
(8 citation statements)
references
References 20 publications
0
8
0
Order By: Relevance
“…In [20] and [21], artificial intelligence was applied to a classification problem to evaluate the voltage stability after large disturbances. To further extract time series features, a spatiotemporal series model representing the short-term voltage evolution trend was proposed in [22,23] to accurately capture the time features of the voltage response and accurately assess the voltage stability [24]. Some studies have developed a data-fault-tolerant method for voltage stability assessment after a fault occurs, which also has high assessment accuracy when data are missing [25].…”
Section: Literature Reviewmentioning
confidence: 99%
“…In [20] and [21], artificial intelligence was applied to a classification problem to evaluate the voltage stability after large disturbances. To further extract time series features, a spatiotemporal series model representing the short-term voltage evolution trend was proposed in [22,23] to accurately capture the time features of the voltage response and accurately assess the voltage stability [24]. Some studies have developed a data-fault-tolerant method for voltage stability assessment after a fault occurs, which also has high assessment accuracy when data are missing [25].…”
Section: Literature Reviewmentioning
confidence: 99%
“…This study employs the temporal ensembling, semisupervised fuzzy c-means clustering (SFCM), k-means clustering algorithm (K-means), and the engineering criteria outlined in [26] to discern four distinct labeled datasets. In this experiment, the SCs of these four labeled datasets are computed and utilized as performance metrics.…”
Section: Performance Testing Of Temporal Ensemblingmentioning
confidence: 99%
“…Reference [25] developed a data-driven STVSA method based on graph convolution network (GCN). Reference [26] introduced an STVSA model based on graphical neural network (GNN), which can quickly identify fast voltage collapses (FVC) and delayed voltage recovery events caused by faults. Reference [27] proposed a real-time STVSA method by combining temporal convolutional neural network (CNN) and LSTM.…”
Section: Introductionmentioning
confidence: 99%
“…Graph neural networks (GNNs) [16], which have emerged in recent years, can model nonEuclidean spatial data and capture the spatial-temporal connections of data, providing a new idea to solve the problem of topology feature extraction. The advantage of the GNN is recognized by some researchers and leveraged in several different applications in power systems, including power system transient stability assessment [17,18], fault location [19], fault classification [20], feeder generation [21], power flow calculation [22], stability control [23] etc. Particularly, graph sample and aggregate (GraphSAGE) [24] is a general inductive framework that can efficiently generate node embeddings for previously unseen nodes by using node feature attribute information.…”
Section: Introductionmentioning
confidence: 99%