Non-contact fault detection of overhead insulators in distribution networks based on sound and light collaboration

Abstract Due to the low detection accuracy of traditional non-contact fault detection methods for overhead insulators in distribution networks, a non-contact fault detection method is proposed for overhead insulators in distribution networks based on sound and light collaboration. This method first uses an acoustic and optical signal acquisition device to capture insulators’ acoustic and optical signals during operation, and preprocesses the collected data to remove noise and improve signal clarity. Next, key fault features are extracted from these signals, such as the analysis of the acoustic frequency spectrum and the identification of spectral features, which help distinguish insulators’ normal working and fault states. Finally, the extracted features are classified using the support vector machine algorithm to achieve non-contact fault detection of overhead insulators in the distribution network. The experimental results show that this method can accurately and quickly detect the faults of overhead insulators in the distribution network, with high detection accuracy and practicality, providing strong technical support for the safe operation of the distribution network.