FUZZY-RANDOM ANALYSIS OF STEEL STRUCTURES

The uncertainty of stochastic computations for which input random quantities are introduced in the subjective manner is presented in this paper. To obtain a correct stochastic computation, it is necessary to have exclusively objective statistical information based on experiments including the correlation matrix at the disposal. If this information is not available, the vague (fuzzy) uncertainty is to be quantified by applying the so-called fuzzy analysis. In the present paper, these problems are demonstrated on the example of an analysis of fuzzy uncertainty of probability assessment of the reliability of a steel member under compression. The output is a fuzzy set of failure probability. The single crisp control output of failure probability was obtained with use of the defuzzification centre of gravity method. The value obtained was by 200 percent higher than that obtained by the classical stochastic computation. In conclusion, the applicability possibilities of the complex fuzzy-random uncertainty analysis at the probability assessment of the reliability of structures by the SBRA method.