Artificial neural network based fault prediction framework for transformers in power systems

Venugopal, K R; Madhusudan, Puttaswamy; Amrutha, A.

doi:10.1109/iccmc.2017.8282519

Cited by 7 publications

(2 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, due to the diversity of power system events, choosing proper thresholds for different scenarios is not an easy task. More recently, purely data-driven classification methods using PMU measurements has begun to gain traction [8], [9]. Supervised machine learning (ML) approaches (see, for example, [10]- [15]) for event identification mainly suffer from weakly labeled historical events.…”

mentioning

confidence: 99%

A Machine Learning Framework for Event Identification via Modal Analysis of PMU Data

Bazargani¹,

Dasarathy²,

Sankar³

et al. 2022

Preprint

View full text Add to dashboard Cite

Power systems are prone to a variety of events (e.g. line trips and generation loss) and real-time identification of such events is crucial in terms of situational awareness, reliability, and security. Using measurements from multiple synchrophasors, i.e., phasor measurement units (PMUs), we propose to identify events by extracting features based on modal dynamics. We combine such traditional physics-based feature extraction methods with machine learning to distinguish different event types. Including all measurement channels at each PMU allows exploiting diverse features but also requires learning classification models over a high-dimensional space. To address this issue, various feature selection methods are implemented to choose the best subset of features. Using the obtained subset of features, we investigate the performance of two well-known classification models, namely, logistic regression (LR) and support vector machines (SVM) to identify generation loss and line trip events in two datasets. The first dataset is obtained from simulated generation loss and line trip events in the Texas 2000-bus synthetic grid. The second is a proprietary dataset with labeled events obtained from a large utility in the USA involving measurements from nearly 500 PMUs. Our results indicate that the proposed framework is promising for identifying the two types of events.

show abstract

mentioning

confidence: 99%

A Machine Learning Framework for Event Identification via Modal Analysis of PMU Data

Bazargani¹,

Dasarathy²,

Sankar³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…There are various types of clustering methods applied to astronomical data such as partition-based methods which search for optimal cluster centers [129]; density-based methods which consider points with high density as the same cluster such as DBSCAN [125], which has recently gained popularity [92,164]; hierarchy-based methods that use bottomup agglomerative grouping of similar samples based on various types of distance computations (e.g., single linkage, average linkage, or full linkage) [107]; model-based methods that mainly use neural network models or probability-based models [118,217]; and graph-based methods that op-erate on graphs [232]. For a full review of clustering methods recently applied to astronomical spectra data, we refer to the study by Yang et al [232].…”

Section: Clustering In Astronomymentioning

confidence: 99%

Visual exploration of high-dimensional data using dimensionality reduction

Kim

View full text Add to dashboard Cite

This Work is dedicated By a humble native of Flatland In the hope that Even as he was initiated into the mysteries Of three dimensions Having been previously conversant With only two So the citizens of that celestial region May aspire yet higher and higher To the secrets of four five or even six dimensions Thereby contributing To the enlargement of the imagination And the possible development Of that most rare and excellent gift of modesty Among the superior races Of solid humanity.-Edwin A. Abbott, Flatland: A Romance of Many Dimensions v * Equal contribution. ix xi contents 2.4 Results 39 2.4.1 Synthetic data: qualitative evaluation 39 2.4.2 Real-world data: qualitative evaluation 42 2.4.2.1 Data sets and their traits 42 2.4.2.2 HD-SDR applied to real-world data sets 43 2.4.3 Quantitative Evaluation 44 2.4.3.1 Evaluation of LGC 46 2.4.3.2 Evaluation of HD-SDR 46 2.5 Application to astronomical data 50 2.6 Discussion 57 2.6.1 Scalability 57 2.6.1.1 Speed 57 2.6.1.2 High-dimensional data 59 2.6.2 Data distortion 60 2.6.3 Relation to clustering 61 2.6.4 Preservation of outliers 61 2.6.5 Limitations and future work 63 2.6.5.1 Undersegmentation and oversegmentation 63 2.6.5.2 Noise-free data 64 2.6.5.3 Parameter setting 64 2.6.5.4Post-processing in dimensionality reduction 65 2.6.5.5LGC for t-SNE and UMAP 66 2.6.5.6Selection of baseline DR method 66 2.7 Conclusion 66

show abstract

Fault Diagnosis and Prediction Method of Distribution Transformer in Active Distribution Network Based on Data Augmentation

Li,

Wang,

Lin

et al. 2023

2023 3rd International Conference on Electrical Engineering and Mechatronics Technology (ICEEMT)

View full text Add to dashboard Cite

Artificial neural network based fault prediction framework for transformers in power systems

Cited by 7 publications

References 10 publications

A Machine Learning Framework for Event Identification via Modal Analysis of PMU Data

A Machine Learning Framework for Event Identification via Modal Analysis of PMU Data

Visual exploration of high-dimensional data using dimensionality reduction

Fault Diagnosis and Prediction Method of Distribution Transformer in Active Distribution Network Based on Data Augmentation

Contact Info

Product

Resources

About