טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentBukh Julia
SubjectFuzzy Logic Implementation Model in Risk Assessment
DepartmentDepartment of Quality Assurance and Reliability
Supervisor Dr. Phineas Dickstein


Abstract

Risk is encountered in almost everyday activities, e.g. crossing the street. In our everyday language risk is - the threat of an undesirable event (injury, damage, or loss). The "objective" evaluation of risk, carried out by specialist, aims of associating a quantitative measure to the probability of a person to suffer an adverse effect from a certain activity over given period of involvement.

Besides the objective assessment of risk, there exists the "subjective" public opinion toward certain activities, which is influenced by perceptive factors such as fear, knowledge, familiarity, trust reversibility of consequence and many other factors related to social, cultural and political backgrounds.

So, the aim of this study is a quantification of public risk perception toward nuclear field. The proposed model includes psychological - voluntariness, familiarity and "numerical" factors influence risk perception, such as probability of occurrence and severity of consequence of nuclear accident. Since the half of these factors can be characterized only by qualitative expressions and the determination all of them are linked with vagueness, imprecision and uncertainty, the most suitable method for the risk level assessment is Fuzzy Logic; due to the fact that, Fuzzy Logic approach using IF-THEN rules can model the qualitative aspects of human knowledge and reasoning processes without employing precise quantitative analyses. This study focuses on the development and representation of linguistic variables to model risk levels associated with factors that influence risk perception. These variables are then quantified using fuzzy sets.

Since, that the factors have not the same influence on the perception of risk, in purpose to represent the importance of each parameter the Analytic Hierarchy Process was employed.