A Method for Identifying Multiple Interference Sources in Complex New Energy Grid Connection Based on Deep Learning Algorithms

Due to the sub synchronous oscillation characteristics of complex new energy grid connection, the stability of power output cannot be guaranteed, resulting in low accuracy in identifying multiple interference sources. Therefore, a complex new energy grid connected multi interference source identification method based on deep learning algorithms is proposed. Using singular value decomposition (SVD) to remove interference source noise, extract interference source features, and obtain complex new energy grid connected multiple interference factors based on artificial intelligence models. Integrating attention mechanisms into deep attention networks to calculate the probability of interference source anomalies, achieving identification of multiple interference sources. The experimental results show that the deep learning algorithm proposed by the research method has very ideal convergence, and the AUC index under the research method is high, with a lower error rate, always remaining below 2%.