Type 1 diabetes is one of the major concerns in current medical studies. Traditional clinical practice involves non-autonomous manual injection of insulin in the blood, while current research in the field of autonomous regulation of blood glucose concentration mostly focuses on model-based control techniques. This paper introduces a novel Reinforcement Learning-based controller for autonomous glycemic regulation in the treatment of type 1 diabetes, building on the Deep Deterministic Policy Gradient algorithm. The proposed control method is validated through in-vitro simulations on the Bergman glucoregulatory model, proving that it successfully preserves healthy values of blood glucose concentration, while overcoming both standard clinical practice and classical model-based control techniques in terms of both control effort and computational efficiency for real-time applications.
Deep Deterministic Policy Gradient Control of Type 1 Diabetes / Baldisseri, Federico; Menegatti, Danilo; Wrona, Andrea. - (2024), pp. 868-873. (Intervento presentato al convegno 2024 European Control Conference (ECC) tenutosi a Stoccolma) [10.23919/ecc64448.2024.10591007].
Deep Deterministic Policy Gradient Control of Type 1 Diabetes
Baldisseri, Federico
;Menegatti, Danilo;Wrona, Andrea
2024
Abstract
Type 1 diabetes is one of the major concerns in current medical studies. Traditional clinical practice involves non-autonomous manual injection of insulin in the blood, while current research in the field of autonomous regulation of blood glucose concentration mostly focuses on model-based control techniques. This paper introduces a novel Reinforcement Learning-based controller for autonomous glycemic regulation in the treatment of type 1 diabetes, building on the Deep Deterministic Policy Gradient algorithm. The proposed control method is validated through in-vitro simulations on the Bergman glucoregulatory model, proving that it successfully preserves healthy values of blood glucose concentration, while overcoming both standard clinical practice and classical model-based control techniques in terms of both control effort and computational efficiency for real-time applications.File | Dimensione | Formato | |
---|---|---|---|
Baldisseri_Deep_2024.pdf
accesso aperto
Note: DOI 10.23919/ECC64448.2024.1059100
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
941.8 kB
Formato
Adobe PDF
|
941.8 kB | Adobe PDF |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.