A Deep Neural Network-Based Multiple Renewable Energy Stations Short Circuit Ratio Prediction Method

With the growing penetration of renewable energy sources such as wind and photovoltaic power, the strength of local power grids weakens, leading to transient overvoltages and oscillations that threaten system stability. Therefore, accurately assessing grid strength has become an urgent necessity. The multiple renewable energy stations short circuit ratio (MRSCR) is a critical indicator for evaluating grid strength, quantifying voltage support at the point of common coupling (PCC) of renewable energy stations (RES) and analyzing integration capacity and instability issues. However, in practical applications, the significant uncertainty in renewable energy output, combined with the periodic updates of MRSCR at RES using offline computational methods, results in outdated monitoring data which hinder real-time security assessments of power systems. To accurately and efficiently assess the impact of renewable energy output uncertainty on grid strength, this paper proposes a deep neural network (DNN)-based prediction method for MRSCR. This paper first establishes a mathematical model for MRSCR based on a simplified equivalent circuit of the sending-end power system containing multiple RES. Then, the DNN is employed for accurate MRSCR prediction, leveraging its strengths in feature extraction and generalization. Finally, the proposed method is tested on a real regional power system in China, and the results validate its effectiveness. Compared to other neural networks, the DNN model offers higher precision in predicting MRSCR with RMSE low as 0.0542 and MAE as 0.0438 in the case study.