APPLICATION OF FUZZY LOGIC FOR MAKING DECISIONS ABOUT EVACUATION IN CASE OF FLOODING

Abstract

We are talking about natural disasters, such as flooding, which can be predicted a few
hours before they occur so that evacuation of the population can be organized. Evacuation
means that people in disaster areas must leave these areas and reach shelters. The article pr esents
an analysis of the decision-making process on evacuation, the main criteria determining
the decision and the main stages of using fuzzy logic to make a decision on evacuation based on
qualitative and quantitative values of the decision-making criteria. These stages include selection
of criteria, determination of qualitative input and output variables, fuzzification of variables,
definition of the base of fuzzy rules, construction of fuzzy inference, visualization of results
and sensitivity analysis. When modeling, the following criteria were taken into account: the
predicted flood level, the level of danger, the vulnerability of the area of the expected flood and
the possibility of safe evacuation. The predicted flood level was based on the parameters of the
maximum level and the rate of water rise. The hazard level reflected the physical characteristics
of the flood and its potential impact on the safety of people in the flood area. The vulnerability
of the area of the expected flood was defined as the inability at the local level to prevent people
from direct contact with flood waters during the event. The possibility of safe evacuation was
defined as a set of limitations and potential negative aspects that could delay or hinder the successful
evacuation. The description of qualitative variable criteria for making a decision on the
need for evacuation, examples of determining the base of fuzzy rules are presented. The fuzzy
model is implemented using Matlab Fuzzy Logic Toolbox. The procedure of fuzzy inference and
interpretation of the solution and a model of several scenarios and flood situations are described.
The method by which a fuzzy model of decision-making on evacuation can be applied in
combination with a geoinformation system is considered. The actions related to the need for
evacuation for various scenarios and circumstances are presented.