Comparison of several data-driven non-linear system identification methods on a simplified glucoregulatory system example

In this study, several advanced data-driven non-linear identification techniques are compared on a specific problem: a simplified glucoregulatory system modelling example. This problem represents a challenge in the development of an artificial pancreas for Type 1 diabetes mellitus treatment, since for this application good non-linear models are needed to design accurate closed-loop controllers to regulate the glucose level in the blood. Block-oriented as well as state-space models are used to describe both the dynamics and the non-linear behaviour of the insulin-glucose system, and the advantages and drawbacks of each method are pointed out. The obtained non-linear models are accurate in simulating the patient's behaviour, and some of them are also sufficiently simple to be considered in the implementation of a model-based controller to develop the artificial pancreas.