Increasing the Quality Indicators of the Functioning of Fuzzy Solvers at the Defuzzification Stage

An approach to the implementation of the defuzzification procedure in microprocessor systems of fuzzy information processing is proposed. A computationally efficient model of the defuzzification operation is considered, based on the use of the concept of the measure of validity of a logical conclusion as a composition of the current values of the reliability of all elements of the conditional part of the rule that proposes this conclusion. The application of the approach provides a significant improvement in the quality of fuzzy approximation with little complexity of algorithmic and hardware-software tools. Examples of the use of this approach in practical problems are given.