The Calculation of Static Voltage Stability Limits Based on Convolutional Neural Networks

In the context of large-scale wind power integration, traditional methods for calculating static voltage stability limits suffer from low computational efficiency and poor adaptability. To address these issues, this paper proposes a fast calculation method for static voltage stability limits based on Convolutional Neural Networks (CNN). The Random Forest algorithm is used to select feature attributes that are strongly correlated with the static stability limits of the transmission sections. These key feature attributes are then input into the CNN for training, establishing a nonlinear mapping relationship between the transmission section limits and the key features. Case studies show that the proposed method significantly improves computation speed compared to traditional continuation power flow algorithms, with prediction errors for static voltage stability limits of less than 5%. This method is suitable for calculating static voltage stability limits of transmission sections in large-scale wind power aggregation systems.