Improving the Accuracy of Fault Identification for Transmission Line Defectsy

The identification rate of transmission line defects and faults is an important assessment indicator for the transformation of the transmission profession towards digital intelligence. After investigation and analysis, the team members found that the low accuracy of intelligent monitoring device channel defect and fault identification is the key factor affecting the identification rate of transmission line defects and faults. In response to the crux, 6 end causes were identified through software and hardware analysis, and 4 main causes were identified through various analyses. In response to the key factors, on-site investigation and testing were carried out one by one, and four strategies were proposed to improve the chip's main frequency and NPU computing power, replace cameras with higher pixels, improve image compression algorithms, and increase the number of sample libraries. After the implementation of the countermeasures, the effectiveness inspection was carried out, and the accuracy of identifying channel defects and faults in the intelligent monitoring device for transmission lines was increased from 86.68% to 91.50%. The crux was successfully resolved, and the activity achieved initial results.