Modeling the measurement uncertainty with Fuzzy approach

Abstract

There are several types of uncertainty in a material characterization arisen from different sources of measurement errors, such as methodological, instrumental, and personal. As a reason of the uncertainty in material models, it is plausible to consider model parameters in an interval instead of a singleton. The probability theory is widely known method used for the consideration of uncertainties by means of a certain distribution function and confidence level concept. In this study, fuzzy logic is considered within a material characterization model to deal with the uncertainty coming from random measurement errors. Data points are treated using fuzzy numbers instead of single values to cover random measurement errors. In this context, an illustrative example, prepared with core strength-rebound hammer data obtained from a concrete structure, is solved and evaluated in detail. Results revealed that there is a potential for fuzzy logic to characterize the uncertainty in a material model arisen from measurement errors.