2020
DOI: 10.1109/access.2020.2986726
|View full text |Cite
|
Sign up to set email alerts
|

Improved Self-Organizing Map Clustering of Power Transformer Dissolved Gas Analysis Using Inputs Pre-Processing

Abstract: Ability to organize data spatially while conserving the topological relation between data features makes the Self Organizing Map (SOM) a very useful tool for analysis and visualization of high dimensional data such as a power transformer's Dissolved Gas Analysis (DGA). Past SOM application required large historical data for its training and has limited fault detection sensitivity. In this paper, the effects of input features and data normalization are studied to enhance SOM's clustering. SOM is trained using D… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
14
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 31 publications
(14 citation statements)
references
References 37 publications
0
14
0
Order By: Relevance
“…PD for partial discharge, D 1 and D 2 for arcing, and T 1 , T 2 and T 3 for thermal overheating. The results obtained are compared with those obtained with IRM [16], DT [10], TRT [18], GT [14] and SOM clusters [33]. Tables 18 and 19 summarize the comparison between proposed diagnostic method and other diagnostic methods obtained with 117 cases of IEC TC10 databases.…”
Section: Validation and Comparison With Other Conventional Methods Us...mentioning
confidence: 99%
“…PD for partial discharge, D 1 and D 2 for arcing, and T 1 , T 2 and T 3 for thermal overheating. The results obtained are compared with those obtained with IRM [16], DT [10], TRT [18], GT [14] and SOM clusters [33]. Tables 18 and 19 summarize the comparison between proposed diagnostic method and other diagnostic methods obtained with 117 cases of IEC TC10 databases.…”
Section: Validation and Comparison With Other Conventional Methods Us...mentioning
confidence: 99%
“…Moreover, as it requires a tremendous cost to perform thorough visual inspection to recognize incipient faults every time, most DGA data in industrial fields is unlabeled. Since sparse, fault-labeled data results in limitations in the ability to confirm reliable quantitative results, additional qualitative methods have been developed, such as high-level feature visualization in 2D space using unsupervised dimension reduction algorithms (e.g., t-stochastic neighbor embedding (t-SNE) and self-organizing map (SOM)) [2,31]. However, it is worth noting that some key information associated with fault diagnosis can be lost during the dimension reduction procedure.…”
Section: B Saat-based Fault Diagnosis Methodsmentioning
confidence: 99%
“…However, it is worth noting that some key information associated with fault diagnosis can be lost during the dimension reduction procedure. Moreover, since both t-SNE and SOM have the ability to cluster the neighboring data, the correlation between high-level features cannot be guaranteed [31,32,46].…”
Section: B Saat-based Fault Diagnosis Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…The main method for diagnosing the power transformers' technical state is analysis of dissolved gases, which result from the oil degradation during power transformer operation. The ratio of certain gases' concentrations allows one to detect the type of equipment damage and its location [9][10][11][12][13]; 2. Power factor analysis.…”
Section: Introductionmentioning
confidence: 99%