Algorithm for assessing quality of fuzzy expert information

The purpose presents the algorithm for software implementation of the expert system block to perform diagnostics and update the parameters of complex technical systems, operating in changing conditions when external factors assessment is presented in the form of fuzzy expert information. The block is intended to input, analyze, and update the expert knowledge, on the basis of which a fuzzy logic inference of values of the system parameters is performed. The accuracy of the inferred parameter significantly depends on the adequacy of the expert information to real conditions, which determines the urgency of generating qualitative expert information. The proposed new algorithm includes identifying all the characteristics of consistency, specifying critical values of consistency indices, it provides feedback with experts for immediate information updating, performs selection of optimal models of linguistic variables and construction of the generalized membership functions with consideration of the hierarchy of expert information. The algorithm application makes the process of expert information generation completely automated. A model example of generating expert information has been provided.