Model Based Control of Blood Glucose for Type 1 Diabetic Patients Using Artificial Pancreas

Abstract

Diabetes Mellitus is a chronic disease affecting millions of people around the world. Most common practise to treat this disease is taking insulin shots or injections that are taken prior to meals. However, this method of taking the drug manually does not necessarily provide required blood glucose control in the body and hence may lead to various other life threatening diseases. With automation in medical technology, an alternate solution is coming into development called artificial pancreas to automatically measure blood glucose in the body and deliver insulin accordingly and therefore maintain tight glucose control. To study this problem, a brief introduction is provided as to justify how this biomedical problem serves to be control system problem. For that various mathematical models and control algorithms are were studied. The aim of this thesis is to design controller for regulation of blood glucose for Type 1 diabetic patients through automated insulin delivery done in-silico using Bergman minimal model. Two controllers are designed for analysis namely, internal model control based PI controller and fractional order controller are designed and exploited in this field in order to maintain normal blood sugar in the following of meal (disturbance).