Regulation of Blood Glucose Level in Patients with Type I Diabetes using Sliding Mode Control

The automatic control system of the blood glucose level is known as an artificial pancreas, whose design and implementation has been researched in recent years. The closed-loop control of the blood glucose concentration deals with some challenges such as sudden decreasing of the blood glucose level, the effects of the hydrocarbon intake of meals, physical exercise and physiological changes due to nervous pressure or stress on the blood glucose level. There are also challenges like day-to-day and patient-to-patient variations. In this paper, a novel adaptive fuzzy integral sliding mode control is presented for the regulation of the blood glucose level. The proposed method is evaluated by simulating nine virtual patients. In the simulations, based on the presented virtual patient model, not only the controller performance in dealing with the effects of the hydrocarbon consuming and physical exercises on the glucose concentration is studied, but also the effects of patient-to-patient variations are investigated simultaneously. For each subsystem of the insulin pump and continuous glucose sensor, a dynamic model is considered. The simulation results and a comparison with the results of the PID control, model predictive control, and adaptive fractional sliding mode control show the efficiency of the proposed method.