Identification of a control-relevant glucose model of type I diabetic patient with physiological meaning for closed-loop strategies in artificial pancreas

RIVADENEIRA, PABLO S.; GONZÁLEZ, ALEJANDRO HERNÁN; GODOY, JOSÉ LUIS

Viena

Conferencia; 11th Advanced Technology and Treatment for Diabetes (ATTD 2018); 2018

ATTD

Minimal models are usually used in the design of control strategies in the so-called device ?artificial pancreas? (AP). Simple black-box minimal models have been obtained from experimental data, but they have limited representation of a T1DM physiology. Also, some gray-box models based on physiological knowledge and real data have been proposed but with equilibria characterizations far from being realistic. Recently, a new model with realistic equilibria was developed in [1], which also permits an explicit computation ofthe tools of the functional insulin therapy (FIT) - IOB, DIA, and others. A further improvement of that is done here, which shows better characteristic from both, real patient physiological description and control relevance point of view.