Research on Secondary Analysis Method of Synchronous Satellite Monitoring Data of Power Grid Wildfire

In the process of monitoring the power grid wildfire, synchronous satellite may produce false fire point information. By using the hot spot monitoring data of Himawari-8 synchronous satellite and field verification information during 2015-2020, this study constructs a secondary analysis algorithm of hot spot monitoring data based on binary decision tree, and analyzes the effectiveness of the algorithm in the improvement of fire identification. The research shows that the algorithm of hot spot secondary analysis has advantages in adapting to data characteristics, fast automatic identification, and autonomous optimization. After training and testing the monitoring data of 2015-2020 hot spots data, the binary decision tree algorithm can significantly reduce the hot spots monitoring error rate of synchronous satellite, with a 100% accuracy rate of the true fire and a 77.1% accuracy rate of the false fire. There is a certain difference between North and South of monitoring fire data of synchronous satellite. As a whole, the accuracy of North fire is higher, and the more south the lower accuracy is.