|
Introducing a Novel Method for Identifying and Classifying Power Quality Disturbances Using LSTM Neural Network, Wavelet Transform, and Intrinsic Mode Decomposition
|
Gholamreza Shahabadi *   , Siavash Eshaghi , Fatemeh Bidar |
| Department of Electrical Engineering, University of Applied Science and Technology, South Khorasan Branch, Iran |
|
|
Abstract: (11 Views) |
In order to address issues related to power quality disturbances, it is necessary to accurately identify type of disturbance. To achieve this, three distinct stages including signal decomposition and analysis, feature selection, and appropriate classification need to be designed. In this paper, a comprehensive set of power quality disturbances has been extracted. An intelligent proposed framework for the correct identification and classification of various power quality disturbances is presented. Signals related to power quality disturbances have been examined separately and in combination. To extract features, a combination of time-frequency wavelet transforms and intrinsic mode decomposition has been utilized. Using this combination and statistical parameters, a 28-length feature vector has been extract. For classification, LSTM neural networks have been employed. Results obtained after training and evaluating the neural network demonstrate the high accuracy of the model.
|
|
| Keywords: Power Quality Disturbance, Wavelet Transform, Empirical Mode Decomposition, LSTM Neural Network |
|
|
Full-Text [PDF 803 kb]
(5 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2024/09/29 | Accepted: 2024/11/30 | Published: 2024/11/30
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
