A wildfire occurrence risk model based on a back-propagation neural network-optimized genetic algorithm

To reduce the impact of wildfires on the operation of power systems, a back-propagation neural network (BPNN) model is used to evaluate the wildfire risk distribution after feature selection. Data from 14 types of wildfire-related features, including anthropogenic, geographical, and meteorological factors, were collected from public data websites and local departments. The weight ranking was calculated using filtering and wrapper methods to form five feature subsets. These are used as the input sets of the BPNN model training, and network parameters are optimized by genetic algorithm (GA). Finally, the optimal feature subset is chosen to establish the optimal BPNN model. With the optimal model, the prediction results are graded to draw a wildfire risk distribution map. Situated in medium-, high-, and very-high-risk zones are 90.26% of new fire incidents, indicating the applicability of the proposed BPNN model.

Automated summary

A back-propagation neural network (BPNN) model is used to evaluate the wildfire risk distribution by selecting the optimal feature subset from 14 types of wildfire-related features. The model is trained with five feature subsets and optimized using genetic algorithm (GA). The prediction results from the optimal model are used to draw a wildfire risk distribution map, showing that 90.26% of new fire incidents occur in medium-, high-, and very-high-risk zones, indicating the practical applicability of the proposed BPNN model.