Modeling and Control of the Glucose-Insulin-Glucagon System in Type I Diabetis Mellitus

Type I diabetes mellitus (DM) is a chronic disease characterized by the inability of the pancreas to produce enough insulin. This results in a high glucose concentration in the blood. External administration of insulin is required to maintain normoglycemia. The accurate amount of insulin and the right time of administration can be obtained from monitoring and practicing model-based control. A model with insulin as a regulatory hormone has already been developed for DM. It has then been extended by considering both insulin and glucagon as regulatory hormones. The model parameters are estimated using non-linear least-square techniques on data from frequently sampled intravenous glucose tolerance tests (FSIGT) from the literature, thus optimizing the model parameters for type I patients. Using this model, a model-predictive controller (MPC) has been designed to maintain normoglycemia in subjects with DM when subjected to disturbances. Finally, a comparative performance analysis between the MPC and conventional PID controller has been carried out. In conclusion, the model-predictive controller is promising for the control of glucose concentration in subjects with type 1 diabetes mellitus when subjected to disturbances.