Towards Hypoglycemia Prediction and Avoidance for Type 1 Diabetic Patients - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

Towards Hypoglycemia Prediction and Avoidance for Type 1 Diabetic Patients

Résumé

The mathematical properties as positivity and positive invariance are helpful for Type 1 Diabetes mellitus (T1DM) system model to formalize the constraints of positivity on glycemia and insulinemia and the positivity of insulin infusion. Nevertheless, they are seldom used in this context. In this paper, the largest Positively Invariant Set (PIS) is computed for the open-loop system (where only a basal insulin rate is infused). Its practical interest is that hypoglycemia is prevented for any trajectory initiated inside the PIS and hypoglycemia is predicted outside the PIS. These results can be used in both open-loop usual basal-bolus therapy or to support any closed-loop control algorithm to avoid hypoglycemia.

Mots clés

Domaines

Automatique
Fichier non déposé

Dates et versions

hal-01592191 , version 1 (22-09-2017)

Identifiants

  • HAL Id : hal-01592191 , version 1

Citer

Taghreed Mohammadridha, Pablo S. Rivadeneira, Michel Cardelli, Nicolas Magdelaine, Claude H. Moog. Towards Hypoglycemia Prediction and Avoidance for Type 1 Diabetic Patients. 56th IEEE Conference on Decision and Control, IEEE, Dec 2017, Melbourne, Australia. pp.4118 - 4123. ⟨hal-01592191⟩
122 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More