Robust servo control of a novel type 1 diabetic model

• Published in: Optimal control applications & methods Vol. 32; no. 2; pp. 215 - 238
• Main Authors: Kovacs, L, Kulcsar, B, Gyorgy, A, Benyo, Z
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.03.2011
• Subjects: algorithm; Algorithms; blood-glucose control; Computer simulation; control design; delivery; Glucose; glucose tracking; insulin sensitivity; intravenous glucose; invivo; Mathematical models; mellitus; mu-synthesis; Nonlinearity; Patients; robust control; Sensitivity analysis; Structured singular values; systems; tolerance; type 1 diabetes; uncertainty; µ-synthesis
• ISSN: 0143-2087; 1099-1514; 1099-1514
• DOI: 10.1002/oca.963

Robust servo control of a model‐based biomedical application is presented in the article. The glucose–insulin control of type 1 diabetic patients is considered to be solved using the results of post‐modern robust control principles. The paper uses a recently published glucose–insulin model and presents the transformation of the model to describe the dynamics of type 1 diabetes mellitus. The nonlinear plant is then linearized at a given steady state point. In order to characterize the uncertainty around the nominal model in frequency domain, a parametric nonlinear model sensitivity analysis is performed using gridding method. The aim of the paper is to underline the viability of the robust servo, linear µ‐control algorithm tested in highly nonlinear closed‐loop simulation environment. Using two degree‐of‐freedom robust controller, the structured singular value of the closed‐loop is designed to fulfill the robust performance requirements and assure glucose level control. Glucose level tracking is ensured under simulated and realistic exogenous meal disturbances. Copyright © 2010 John Wiley & Sons, Ltd.