Deep learning approach for classification of PQ disturbances

The fast upgrade of the grid, development of new technologies and expanded penetration of renewable resources in the power network result in various power quality issues. This raises the question of how the power quality issues will affect the end point users. This is attracting considerable attention between researchers to investigate and find a suitable approach for precise detection and classification of power disturbances. This paper proposes a deep convolutional neural network for the detection and classification of power disturbances. The model automates the process of feature extraction and classification and establishes a better connection between the feature extractor and classifier. The obtained results have shown high accuracy of the model on the synthetically generated signals. Additionally, the comparison with other proposed models in the literature is presented.