The exogenous insulin needs of a type-1 diabetes patient are crucial to avoid physical complications like hypoglycemia (blood glucose \({>}\) 180 mg/dl) or hyperglycemia (blood glucose \({<}\) 70 mg/dl). The physiological parameters and different states related to the system for regulating blood sugar in the human body are uncertain in a real scenario. As a result, adaptive parameter control logic will be the most effective way to keep blood sugar levels in type-1 diabetic patients under control. In this regard, an adaptive higher order sliding mode control strategy is proposed for controlling blood sugar in type-1 diabetic patients. A second-level multi-model parametric adaptation law has been developed to estimate different parameters of a type-1 diabetic model under parametric uncertainties. To avoid costly sensors, an extended Kalman filter state estimator is used to estimate the unavailable states of the model. With the proposed control scheme, different unavailable states and physiological parameters of the type-1 diabetes mellitus model are estimated, and the glucose level is controlled within a safe range (70.0–180.0 mg/dL) within 120 min. The realization of the planned adaptive control plan is examined for scheduled and unscheduled meal disturbances, high meal problems, and measurement noise, and it performed efficiently for all cases.
