Data Driven Identification of IDDM Patient Model

ARPITA, BHATTACHARJEE and ASHOKE, SUTRADHAR (2014) Data Driven Identification of IDDM Patient Model. In: International Conference on Advances In Computing, Electronics and Electrical Technology CEET 2014, 02 - 03 August, 2014, Kuala Lumpur, Malaysia.

20150609_050121.pdf - Published Version

Download (805kB) | Preview
Official URL:


Prerequisite to the better living of an insulin dependent diabetes mellitus (IDDM) or type-1 diabetic patients is the closed loop blood glucose regulation via subcutaneous insulin infusion and continuous glucose monitoring system (SC-SC route). Closed loop control for blood glucose level in a diabetic patient necessarily uses an explicit model of the process. A fixed parameter full order or reduced order model does not characterize the inter-patient and intra-patient parameter variability. This paper deals with a real time implementation of online identification of frequency domain kernels from the input output data of an IDDM patient. The data-driven model of the patient is identified in real time by solving Volterra kernels up to second order using adaptive recursive least square (ARLS) algorithm with a short memory length of M=2. The frequency domain kernels, or the Volterra transfer function (VTF) are computed by taking the FFTs on respective time domain kernels for a specific length of extended input vector. The dynamic glucose-insulin process model of a IDDM patient in SC-SC route based on the work of Dalla Man et. al. has been constructed in hardware platform that acts as a virtual patient. The validation results have shown good fit of responses with nominal patient in simulation as well as with online identification.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: diabetes mellitus, identification, nonparametric model, Volterra kernels, hardware realization.
Depositing User: Mr. John Steve
Date Deposited: 27 May 2019 12:12
Last Modified: 27 May 2019 12:12

Actions (login required)

View Item View Item