Fuzzy rules by using fuzzy SRA tool: an application to risk analysis for the Vesuvian area

We experiment the use of the tool “Fuzzy SRA” in order to get the most relevant rule in a set of fuzzy rules related to a decision making problem in spatial analysis. The basic idea consists in the usage of the algebraic structure and the related string operations defined by A.Gisolfi and V.Loia (International Journal of Approximate Reasoning, 13, pp. 151-183, 1995) on a domain formed from fuzzy rules: the string attributes are the variables that appear in each logical clause of the antecedent rules. The rule to be adopted is represented in the final string from a triangular fuzzy number with the best central value. This rule will be the most “reliable” rule by considering the observable data defined as “facts”. We test this approach by considering the problem of the eruption risk coming from the famous vulcan Vesuvio. We have considered a rule-set formed from 22 fuzzy rules used to evaluate the impact produced on the exposed values from an eruption risk in each of the 18 Municipal towns located in the so-called “Red Zone”, which is the most exposed zone to the eruption risk, in accordance to the decisions of the Italian Civil Protection Agency. We have considered in the rules the following parameters: seismic vulnerability, urban environment, resident population, archaeological and architectural heritages. Here the “facts” coincide with the data measured from Italian Civil Protection Agency and ISTAT (Italian Statistical Institute), hence we determine the most correspondent rule to these facts.