Design And Implementation Of Forest Fire Probability Prediction System

Forest is one of the most important resources in China,and forest fires is one of the major calamities in forest development.Forest fires not only directly affect the balance of forest ecological,resulting in the loss of economic and ecological resources,but also endanger people’s lives and property.Therefore,developing a effective and practical forest fire prediction system,is a natural approach to reduce the loss caused by forest fires.Forest fire prediction involves many influence factors,complex occurrence mechanism,and unstructured input data.To address these issues,a practical forest fire probability prediction system is designed and implemented.The main work includes:1.Design of the system structure.The system inputs meteorological,geographical,vegetation and population distribution data,computes the probability of forest firevia Bayesian network(BN),outputs the probability and risk level of forest fire.2.Design and implementation of the BN inference system,which builds BN modelwith historical forest fires records,performs probabilistic inference via the junctiontree algorithm.Experimental results on real records of forest fires in Yunnan province and comparison with the Canadian Forest Fire Weather Index(FWI)System demonstrate that the system is feasible and practical.