Mathematical Modeling of Diabetic Patient Model Using Intelligent Control Techniques

Abstract

Diabetes mellitus (DM) is the most common chronic disease, which is categorized into two types: they are type 1 and type 2. Category of type 1 diabetic patients experiences fluctuations in blood glucose (BG) levels that are mainly caused because the pancreas failing to secrete insulin which results in the cause of hyperglycemia or a high increase in blood glucose (BG) level exceeding 150 mg/dl. In such conditions, a patient needs continuous insulin injections throughout their life span. Considering the metabolism of diabetic patients is very complex and nonlinear, the paper mainly focuses on dynamic simulations of glucose–insulin interaction to obtain a new method for regulating blood glucose levels in diabetic patients. Therefore, a model is developed to describe the process of insulin–glucose interaction which is to simulate the system using state-space analysis. The state-space model development is done with the help of the classic linearization method followed by a closed-loop simulation. In this paperwork, two different models such as Lehmann and Sorensen-based human patient models are proposed to regulate the blood glucose levels of diabetic patients, and a comparison is done with the analysis of a mathematical model on both types of proposed models under the category of type 1 diabetic patient models. Later, the models are validated by the help of machine learning algorithms using Simulink software.