AUTHOR=Zhang Hao , Liu Hui , Ma Guoqing , Zhang Yang , Yao Jinxia , Gu Chao 

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

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 10 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2022.1031762

DOI=10.3389/fenrg.2022.1031762

ISSN=2296-598X

ABSTRACT=To reduce the impact of wildfire on the operation of power system, a Back Propagation Neural Network (BPNN) model is used to evaluate the wildfire risk distribution after feature selection. First, the data of 14 types of wildfire-related features, including anthropogenic, geographical, and meteorological factors, are collected from public data website and local department. The weight ranking is calculated using filtering and wrapper methods to form five feature subsets. These feature subsets 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 wildfire risk distribution map. 90.26% of new happened fire incidents were situated in medium, high-risk and very-high-risk zones, indicating the applicability of proposed BPNN model.