Robust Adaptive Identification of Fuzzy Systems with Uncertain Data

This study presents a method of adaptive identification of parameters describing Sugeno fuzzy inference system in presence of bounded disturbances while maintaining the readability and interpretability of the fuzzy model during and after identification. This method do not require any a priori knowledge of a bound on the disturbance and noise and of a bound on the unknown parameters values. The method can be used for the robust and adaptive identification of slowly time varying nonlinear systems using fuzzy inference systems. The suggested method was used to build a fuzzy expert system that approximates the functional relationship between physical fitness and some of the measurable physiological parameters by their real measurements and opinion (human-experiences) of a medical expert.     

OXYGEN QUENCHING IN MICELLAR SOLUTION AND ITS EFFECT ON THE PHOTOCYCLIZATION OF N-METHYLDIPHENYLAMINE

Abstract— Oxygen effects on the photocyclization of N-methyldiphenylamine to N-methylcarbazole were investigated in n-hexane, water, and aqueous surfactant solutions by steady state irradiations and flash photolysis measurements. The reaction sequence in micelles was found to involve the same intermediate steps as in homogeneous solutions. In aerated micellar solutions, the quantum yield of N-methylcarbazole is significantly higher than in n-hexane, while the rate constants of the unimolecular reaction steps show no solvent dependence. The bimolecular dehydrogenation of the intermediate 4a, 4b-dihydro-N-methylcarbazole by oxygen is enhanced in aqueous and micellar solutions, whereas the quenching rate of triplet intermediates by oxygen was not affected. The lesuhs are interpreted using a dispersed phase model of micellar solutions. Special 'micellar effects' need not be invoked since the dependence of the quantum yield on the solvent is shown to be due to the difference in the overall oxygen concentration.