Research on intelligent identification method of lightning arrester deterioration state based on GASFCNN

In recent years, monitoring and identifying the deterioration state of surge arresters has become essential for enhancing the reliability of electrical power systems. This paper introduces a novel method for recognizing the deterioration state of surge arresters, employing Gramian Angular Summation Fields (GASF) and Convolutional Neural Networks (CNN). Initially, GASF technology is utilized to transform the historical operational data of surge arresters into image format, capturing the texture and structural features that characterize the deterioration state. Subsequently, a specially designed CNN model processes these images to extract features and perform classification, thus identifying the deterioration states accurately. The results of this research offer a new approach for the online recognition of deterioration states in critical protective devices on transmission lines, providing significant insights for the safe and reliable operation of power systems.