Power quality disturbance identification using hybrid deep learning in renewable energy systems.

The use of renewable sources of energy has increased significantly in the recent years. These are hybrid wind-solar photovoltaic (Wind-SPV) plants. To ensure that these systems are reliable and customers happy, we should ensure that the quality of power remains high. In this paper, I am going to present a new hybrid deep learning-based approach to the specific diagnosis of power quality disturbances (PQDs) in Wind-SPV integrated networks. The proposed system combines scalograms obtained using the continuous wavelet transform (CWT) with deep neural networks (ResNet and VGG-Net), neighborhood component analysis (NCA), and support vector machine (SVM) classification. MATLAB/Simulink simulations have been carried out with customised IEEE 9-bus test systems and IEEE 13-bus test systems. The proposed method achieved 98.54% accuracy, 97.57% precision, and 95.21 recall on IEEE 9-bus system and 97.17% accuracy, 96.42% precision and 94.08% recall on the IEEE 13-bus system. The results are significantly improved compared to the conventional PQD sensing techniques such as Fourier transform, short-time Fourier transform, and discrete wavelet transform, which can be either fixed resolution-based or have high computing power requirements. The findings demonstrate that the proposed approach is effective in the real-world setting of a hybrid renewable framework, which contributes to the enhancement of the quality of power and aids the global transition to renewed energy.