Blood Glucose Regulation for Type 1 Diabetic Patients at Intensive Care Unit via Continuous Sliding-Mode Algorithms
Résumé
This chapter proposes five continuous sliding-mode control algorithms to regulate blood glucose levels in critically ill patients affected with Type 1 Diabetes Mellitus. All the controllers deal with uncertainties from intrapatient and interpatient variability and external disturbances such as food intake. The proposed schemes do not require meal announcements or patient individualization. The blood glucose measurement and insulin infusion are intravenous. The control objective is to regulate the blood glucose in a normoglycemia range, i.e., 70–180 (mg/dl). The approach is validated in the UVA/Padova metabolic simulator for ten in silico adult patients. The results show excellent performance and minimal risk of hyperglycemia and hypoglycemia events for every sliding-mode algorithm. A quantitative analysis is carried out to shed light on the controller’s performance. To this aim, we consider several indexes regarding blood glucose regulation and the control signal energy. In addition, simulations using a standard (open-loop) protocol used at ICU highlight the workload alleviation obtained with any SMC.