Abstract

The complexity and nonlinearity of the insulin-glucose regulatory system in humans demands for an inverse modeling technique. In this study the insulin-glucose regulatory system was modeled with the help of an inverse feedback fuzzy state space modeling (FFSSM) approach. The feedback insulin regulatory system was presented as a state space model, its input/output parameters fuzzified and the optimized values deffuzzified using the “Optimized Defuzzified Value Theorem”. In order to get a more detailed insight of the system, the properties concerning state stability of the model were studied. The model was tested and analyzed for insulin sensitive and insulin resistant subjects. The model was found state stable for an intercellular glucose level above 3.50080 mgdl^-1 and 20.0080 mgdl^-1 for insulin sensitive and insulin resistive subjects, respectively. At the glucose level lower than this specified level, the insulin-glucose system may stop working and the subject be encountered with some serious fatality. Decreasing the glucose intake below 103.6 and 64.5 mgdl^-1, in case of insulin sensitive and insulin resistant subjects, respectively, may associate unpredictable parameter values to fuzzy parameters depicting low glucose levels.

Key words: Insulin-glucose regulations, feedback systems, fuzzy state space modeling (FSSM), inverse modeling.