Research Specifications

Home \Noise-resilient deep ...
Title
Noise-resilient deep learning-based CNN for static voltage stability analysis in power systems with wind farm integration
Type of Research Article
Keywords
Static voltage stability assessment · Deep learning · Convolutional neural networks · Wind energy · Uncertainty
Abstract
The growing incorporation of renewable energy sources (RESs) into modern power systems has resulted in an increase in the level of uncertainty that is present during network operations. Increases in both uncertainty and load levels put system stability at risk. Since traditional power system analysis approaches are mainly based on physical modeling and numerical computations, they are time-consuming and computationally intensive. This paper presents an approach for assessing the stability of power systems that have a significant amount of RESs. The approach utilizes a deep learning-based convolutional neural network (CNN) and takes into account various parameters such as bus voltage, current, reactive power of the line, and the fast voltage stability index (FVSI). In this particular investigation, the utilization of CNN is a significant step toward enhancing accuracy and improving robustness against the influence of noise on predictions and efficiency. The proposed model, with a high degree of accuracy equal to 99.63% able to assess a wide variety of input patterns and provide predictions on the stability state of the system. Finally, the numerical results of applying the suggested technique to the IEEE 39-bus and 118-bus test systems illustrate the accuracy and quickness of CNN. These findings are compared to the results of other algorithms.
Researchers (First Researcher)، (Second Researcher)، Mahmoud A. Younis (Third Researcher)، Anas Quteishat (Fourth Researcher)، Amin Safari (Fifth Researcher)، Javad Salehi (Not In First Six Researchers)