An automatic insulin infusion system based on LQG control technique

Abstract

This paper makes an attempt to develop clinical simulation model of glucose–insulin interaction and to design an optimal controller to regulate the Blood Glucose (BG) level in Insulin–Dependent Diabetes Mellitus (IDDM) patients. For enhancing the quality of life of the patient, an automated insulin delivery system based on Linear Quadratic Gaussian (LQG) control algorithm is suggested, and to justify its efficacy a comparative analysis with conventional Proportional–Integral–Derivative (PID) control tuned using the Ziegler–Nichols method and optimal H∞ control based on solutions of Riccati equations is presented. For designing of the controller, a ninth–order linearised state–space model of the glucose–insulin interaction process of an IDDM patient has been used. The controller performances are assessed in terms of ability to track a normoglycaemic set point of 81 mg/dL (4.5 mmol/L) in the presence of Gaussian and stochastic noise.