The phenomenon of cable discharge is one of the common fault issues in power systems,and accurately identif-ying discharge current signals helps prevent major power accidents in advance.This paper proposes a hybrid neural net-work model based on Convolutional Neural Networks(CNN)and Gated Recurrent Units(GRU)for the identification of cable discharge current signals.The CNN is employed to extract spatiotemporal features from the discharge signals,while the GRU is used to process the temporal sequence information,enabling classification of different types of discharge sig-nals.Experimental results show that this method outperforms traditional discharge signal processing techniques in terms of accuracy and robustness.
维普
万方数据
