Estimating Combustible Load and Fire Risk in Zhejiang Power Grid

The power grid system is a cornerstone of modern socio-economic development and is crucial for the stable operation of Zhejiang Province, an economically prosperous region. This study tackles the fire risk management challenges in the province ’ s transmission corridors by utilizing remote sensing technology and machine learning algorithms. Specifically, it develops methods for estimating combustible material loads and assessing fire risks. Using high-resolution optical imagery, hyperspectral data, and Synthetic Aperture Radar (SAR) data, along with ground plot surveys, we extracted spectral characteristics, texture features, and structural information from vegetation to create accurate models for tree species classification and combustible material load estimation. This led to the creation of a combustible material distribution map for the transmission corridors throughout the province. In parallel, we integrated meteorological data and regional phenological characteristics to develop a dynamic fire risk assessment model, which resulted in the generation of fire risk level distribution maps. This research significantly improves the accuracy of vegetation monitoring and fire risk assessment in transmission corridors, providing optimized tools for safe power grid management and enhancing inspection and early warning systems. The findings offer a scientific foundation for forest fire prevention and ecological protection strategies, with wide potential for application and dissemination.