Modified Architecture of an Intuitionistic Fuzzy Neural Network and its Application to a Blood Glucose Regulation for Diabetic Patients

In this paper, a modified intuitionistic fuzzy neural network (MIFNN) is introduced to deal properly with large amount external disturbances, linguistic uncertainties and model uncertainties in real-world implementations. The terminal fuzzification is obtained by incorporating membership functions and non-membership functions with the degree of hesitance. The proposed MIFNN is applied to regulate the blood glucose concentration for diabetic patients.