A fuzzy inference rule-based system for the estimation of forest fire risk: the case of Greece

摘要

The purpose of this manuscript is the design, implementation and application of an intelligent fuzzy inference system that estimates the degree of risk due to natural hazards. More specifically, the system aims in the determination of the forest fire risk index corresponding to each Creek forest department. The obtained results can help the decision makers to characterize the areas under study, based on their degree of risk. This effort can provide important aid towards the proper distribution of fire fighting resources. The system provides the decision maker a ranking of the cases (in descending order), according to the magnitude of their risk. It estimates two distinct indices, namely the man caused risk and the natural caused one. The rule-based system has been implemented in Matlab and it uses data vectors stored in MS Excel. The fuzzy membership functions employed are the triangular and the semi-triangular, while all the aggregation, defuzzyfication and implication functions are provided by Matlab's integrated Fuzzy Logic Toolbox.