Blood Glucose Regulation in Patients with Type 1 Diabetes Using Model Predictive Control and Data Reconciliation

Abstract

The need to have mechanisms and technologies for the control of blood glucose levels is essential for people who have diabetes of any type. In this way, this paper presents the results obtained of blood glucose control using Model Predictive Control (MPC) and digital Proportional Integral (PI) control, which has been simulated with virtual patients. Initially, a model describing the dynamics of glucose metabolism in the human body was elaborated and, to make the simulation more real, it was considered the action of disturbances and noises. After modeling the system, the controllers were developed. Signal processing and data reconciliation was done using the Kalman Filter. With the objective of improving the performance of the controllers, the feedback of reconciled data technique was used and the performance of the MPC controller was improved by 83.9% based on the elaborated cost function.