An Ensemble Machine Learning Based Fault Classification Method for Faults During Power Swing

Patil, Dinesh; Naidu, O. D.; Yalla, Preetham; Hida, Satoshi

doi:10.1109/isgt-asia.2019.8881359

Cited by 9 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nowadays, great efforts have been made in developing new methodologies, from both data and model perspectives, for fault detection and isolation (FDI) (Chen and Patton, 2012;Costamagna et al, 2015;Jan et al, 2017;Zhang et al, 2016;Alhelou et al, 2018). Data-driven FDI has received significant attention recently; for example, approaches using deep learning techniques such as long short-term memory (LSTM) networks and convolutional neural networks (CNNs) are popular among the deep neural networks (Chen et al, 2017;Zhang et al, 2017;Patil et al, 2019;Paul and Mohanty, 2019;Qu et al, 2020). An LSTM network (Hochreiter and Schmidhuber, 1997) has advantages for learning sequences containing both short-and long-term patterns from time series (Malhotra et al, 2015), while CNN is a commodity in the computer vision field that is capable of achieving record-breaking results on highly challenging image datasets (Krizhevsky et al, 2012;Zhu et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

Power System Event Classification and Localization Using a Convolutional Neural Network

et al. 2020

View full text Add to dashboard Cite

Detection and timely identification of power system disturbances are essential for situation awareness and reliable electricity grid operation. Because records of actual events in the system are limited, ensemble simulation-based events are needed to provide adequate data for building event-detection models through deep learning; e.g., a convolutional neural network (CNN). An ensemble numerical simulation-based training data set have been generated through dynamic simulations performed on the Polish system with various types of faults in different locations. Such data augmentation is proven to be able to provide adequate data for deep learning. The synchronous generators’ frequency signals are used and encoded into images for developing and evaluating CNN models for classification of fault types and locations. With a time-domain stacked image set as the benchmark, two different time-series encoding approaches, i.e., wavelet decomposition-based frequency-domain stacking and polar coordinate system-based Gramian Angular Field (GAF) stacking, are also adopted to evaluate and compare the CNN model performance and applicability. The various encoding approaches are suitable for different fault types and spatial zonation. With optimized settings of the developed CNN models, the classification and localization accuracies can go beyond 84 and 91%, respectively.

show abstract