Application of Multi-Source Satellite Monitoring Data in Hidden Danger Diagnosis of Transmission Lines

Traditional manual inspection, unmanned aerial vehicle inspection and ground monitoring equipment have some problems, such as low inspection efficiency, large environmental restrictions and incomplete data acquisition. In this paper, the application of multi-source satellite monitoring data in transmission line hidden danger diagnosis is studied. By comprehensively utilizing the complementary advantages of optical satellite Landsat 8 and radar satellite Sentinel-1, the accuracy and efficiency of hidden danger monitoring are improved. Preprocessing of satellite data was performed using methods such as radiometric calibration, atmospheric correction, and geometric correction to extract key features including the Normalized Difference Vegetation Index (NDVI) and surface deformation. An efficient hidden hazard diagnosis model was constructed by combining a Support Vector Machine (SVM) classifier with a Particle Swarm Optimization (PSO) algorithm. The experimental results show that the model has high accuracy and recall rate in identifying hidden dangers of transmission lines, and can accurately identify hidden dangers such as tree barriers, building intrusion and foreign bodies. Under the optimal parameter combination, the accuracy and recall of the model reach 0.938 and 0.915, respectively. The research in this paper provides theoretical basis and technical support for the application of multi-source satellite monitoring data in the diagnosis of hidden dangers of transmission lines, which is helpful to improve the safe operation level of power grid.