Parameter identification and state estimation for a diabetic glucose‐insulin model via an adaptive observer

Roberto Franco (1) , Héctor Ríos (1) , Alejandra Ferreira de Loza , Louis Cassany (2) , David Gucik-Derigny (2) , Jérôme Cieslak (2) , David Henry (2)

1 TecNM - Tecnológico Nacional de México
2 IMS - Laboratoire de l'intégration, du matériau au système

In this article, an adaptive observer is designed for patients with Type 1 DiabetesMellitus. The adaptive observer, synthesized using the so-called Bergman’s Minimal Model, simultaneously estimates the states and the parameter corresponding to the insulin-independent glucose disappearance rate. The adaptive observer dealswith parameter uncertainties,whereas the food intake is regarded as an external disturbance. The adaptive observer relies on intravenous glucose measurements. The state estimation error converges to a neighborhood of the origin despite the effects of the external disturbances and uncertainties, while the parameter estimation error converges in a fixed time to a neighborhood of the origin. The adaptive observer synthesis is given by a constructive method based on linear matrix inequalities. Simulation results show the feasibility of the proposed scheme. Moreover, the approach is validated in UVA/Padova metabolic simulator for ten in silico adult patients.

Mots clés: Adaptive observers; Diabetes Mellitus; Sliding Mode

Domaines: Automatique; Endocrinologie et métabolisme

Roberto Franco, Héctor Ríos, Alejandra Ferreira de Loza, Louis Cassany, David Gucik-Derigny, et al.. Parameter identification and state estimation for a diabetic glucose‐insulin model via an adaptive observer. International Journal of Robust and Nonlinear Control, 2023, 33 (9), pp.5087-5104. ⟨10.1002/rnc.6030⟩. ⟨hal-03543128⟩