OPTIMAL BLOOD GLUCOSE CONTROL IN TYPE 1 DIABETIC PATIENTS BY USING PONTRYAGIN’S MINIMUM PRINCIPLE AND EXTENDED KALMAN FILTER

Artificial Pancreas (AP) is an advanced diabetes management technology. AP requires an automatic control algorithm to determine the level of insulin injection based on the glucose level calculated on the Continuous Glucose Monitoring (CGM) sensor. Bergman Minimal Model (BMM) is a basic model in describing the dynamics of glucose-insulin in the human body. This study aims to determine the optimal control using Pontryagin’s Minimum Principle (PMP), which is subject to reducing glucose levels in type 1 diabetes patients to be within the normal glucose level interval of 80-120 mg/dL. The BMM parameter values will be estimated using EKF to support the acquisition of precise and personal numerical simulations. Based on the control de sign simulation that has been obtained, the optimal control of insulin injection is given maximally (25 mU/L) during the first two hours of observation; then, the level decreases slowly until it reaches 0 at 12 hours of observation. This scenario successfully reduces the patient’s glucose levels at the end of the observation period from 170.4 mg/dL (with out control) to 121.2 mg/dL. This result providing a confident basis for initial future research and development in diabetes management.