Krasnopoyasovskiy A. The information analysis and syntheses of fuzzy controller of the learning control systems

The thesis is performed for the purpose of increasing of functional efficiency of the learning intelligence control system of technical and technological objects and processes in the frame-sets of the information-extreme method of functional-statistical tests (MFST), that is suggested by author and founded on direct estimation of the information ability of the system at the condition of a priori uncertainty and limited number of tests. In the framesets of MFST the hierarchy of categorical models and algorithms of automatic classification, optimizing of spatio-temporal functional parameters of learning control systems in fuzzy controller sets for the purpose of maximization of full asymptotic reliability of control decision, are developed. The questions of estimation of the functional efficiency of learning intelligence control system, purposeful transformation fuzzy decomposition of the space of signs on recognition classes in exact decomposition during learning process are considered and new tools ofinformation technology of analysis and synthesis of learning control system in fuzzy controller sets are devel-oped.