Stability Study of a Neuro-Fuzzy Output System Based on Ratio Area Method

Purpose of research is to study the hypothesis about possibility changing type transition process during training in a neuro-fuzzy inference system based on area ratio method, and properties weight coefficient influence its stability. Methods. An apparatus fuzzy logic used for development output system. At same time, input variables are described by triangular membership functions. Mamdani implication model was compositional rule. A linear defasification. The reverse error propagation method training. Results. In experimental studies, it found that proposed allows change process, namely, transform oscillatory into an aperiodic (monotonic) process. also stability more influenced determined calculating total Thus, obtained results prove: first, neuro-odd ensures transformation transfer characteristics, second, given range characteristics. Conclusion: architecture adaptive described. distinctive feature use odd accessory functions at inputs outputs. Analysis simulation showed s important ensure It necessary establish permissible values coefficient, numerical which turn affect characteristics