The forest fire prediction in JiangXi province based on PSO-GA-BP neural network
Zhijian Yin, Fan Wang, Tiantian Tang, Qiang Luo, Kun Xiang · 2016
In this paper, we collected the meteorological data and forest fire danger rating data of four stations in JiangXi (NanChang, JingDeZhen, GiAn, GanZhou) from 2013 to 2015, and build a neural network fire prediction model.Then use GA, PSO and PSO-GA hybrid algorithm to improve BP neural network.By contrasting the prediction results of BP network, GA-BP network, PSO-BP network and PSO-GA-BP network, the prediction accuracy of PSO-GA-BP network is the highest.The result of experiment shows that the effect of BP network optimized by PSO-GA is the best, compared with GA and PSO.