Control of T1DM via tensor product-based framework

摘要

With this study our goal was to investigate and prove the usability of the Tensor Product (TP)-transformation based modeling and control regarding the control of Diabetes Mellitus (DM). In details, we examined the TP-based modeling possibilities of Type 1 DM (T1DM) by using the simple well-known Minimal Model. We provided a TP-based controller design solution, where the main focus was on the disturbance rejection. Generally, the TP-based framework using the Parallel Distributed Control (PDC)- and Linear Parameter Varying (LPV) theorems via well-conditioned Linear Matrix Inequalities (LMI) to realize the TP-based controller. In this manner the basis of our controller design procedure was the development of the control oriented qLPV model form, the selection and application of appropriate LMIs through which the PDC-based TP controller can be developed and the realization of the control environment. We prove the usability of the developed control solution on the mentioned T1DM model beside unfavorable disturbances.