Lightning trip prediction of overhead transmission lines based on CNN-GRU network

Abstract

Aiming at the high randomness and uncertainty of lightning trips in overhead transmission lines, the prediction accuracy and efficiency of lightning trips are low, and a prediction method of the lightning trip for overhead transmission lines based on CNN-GRU is proposed. Firstly, the feature set of lightning trip prediction influencing factors of overhead transmission lines is constructed based on the main characteristics influencing factors of overhead transmission lines and operating environment characteristics influencing factors. Secondly, the association rules are used to quantify the correlation between influencing factors and lightning trips. Finally, the Convolutional Neural Network (CNN) -gated Recurrent Unit (GRU) combined network is used to extract the high-dimensional internal connections between the influencing feature factors and line lightning tripping and train them. Combined with practical examples, the prediction results show that the CNN-URU prediction model proposed in this paper has higher prediction accuracy and efficiency than other prediction models.