Algoritmus založený na hladké optimalizaci a jeho použití při řešení nehladkého optimalizačního problému, jenž vzniká při maximalizaci kriteriálního funkcionálu s fuzzy vstupními daty

In maximum range problems, the input parameters of an analyzed model are uncertain and the range of the model output values is searched for. If the input values form a fuzzy set with a known membership function, then the goal is to determine the membership function of the output values, that is, to determine the fuzziness of the quantity of interest. A recently proposed integration-based input membership function and the common .alpha.- cut technique are used to achieve this goal. Although this approach can lead to a nonsmooth optimization problem, it is shown that the nonsmooth problem can be decomposed into a sequence of smooth subproblems and solved by standard smooth optimization routines