An adaptive and personalized multivariable artificial pancreas system is proposed for effective glycemic control and disturbance rejection... Show moreAn adaptive and personalized multivariable artificial pancreas system is proposed for effective glycemic control and disturbance rejection without manual user announcements for meals and exercise. Adaptive models identified through system identification techniques are integrated with a physiological compartment model to characterize the time-varying glucose-insulin dynamics. The real-time estimation of plasma insulin concentration to quantify the insulin in the bloodstream in patients with type 1 diabetes mellitus is presented. The identified time-varying models are employed for the design of an adaptive model predictive control formulation that is cognizant of the plasma insulin concentration. A feature extraction method based on glucose measurements is used to detect rapid deviations from the desired set-point caused by significant disturbances and subsequently modify the constraints of the optimization problem for negotiating between the aggressiveness and robustness of the controller to suggest the required amount of insulin. A predictive hypoglycemia module with carbohydrate suggestion is also designed to prevent any potential hypoglycemia events. A controller performance assessment algorithm is developed to analyze the closed-loop behavior and modify the parameters of the artificial pancreas control system. To this end, various performance indices are defined to quantitatively evaluate the controller efficacy in real-time. The controller assessment and modification module also incorporates on-line learning from historical data to anticipate impending disturbances and proactively counteract their effects. Show less
