A microgrid is a network consisting of one or several loads and distributed generation (DG) sources that operate as a single aggregate load or source. They have become an increasingly popular method to manage DG sources at the distribution level. As such, fault detection and protection of microgrids is a topic of current research. In this work, we study the application of deep learning based techniques such as Long Short Term Memory Networks (LSTM) to classify faults, since accurate real time fault classification can support superior grid operation and prevent relays from closing non-faulted phases. Here, we used Matlab/Simulink to model a simple microgrid and simulate various types of faults to generate training data for the neural network model. Additive White Gaussian Noise (AWGN) and Additive Impulsive Gaussian Noise (AIGN) are added to the signal to mimic real world data. We used Discrete Wavelet Transform (DWT) and Multi-Resolution Analysis (MRA) to process the simulated signals, which are subsequently used to train a LSTM to classify eleven types of faults. We investigate the application of Convolutional Neural Networks (CNN) for fault classification by transforming the simulated signals into images, which are used to train the CNN to classify the same faults. The investigation showed that the LSTM out-performed the CNN and achieved high accuracy in classifying the faults using half the data.
|Commitee:||Talebi, Mohammed, Kwon, Seok-Chul|
|School:||California State University, Long Beach|
|School Location:||United States -- California|
|Source:||MAI 82/4(E), Masters Abstracts International|
|Subjects:||Electrical engineering, Artificial intelligence, Systems science|
|Keywords:||Classification, Convolutional Neural Networks, Deep Learning, LSTM Neural Networks, Microgrids, Power System Faults|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be