Hardware design for blood glucose control based on the Sorensen diabetic patient model using a robust evolving...

Area of Science:

• Biomedical Engineering
• Control Systems Engineering
• Computational Intelligence

Background:

• Diabetes Mellitus (DM) poses significant public health challenges, necessitating advanced engineering solutions to prevent complications.
• Existing models often lack the ability to account for dynamic patient factors like stress, meals, and exercise, impacting glucose regulation.

Purpose of the Study:

• To develop a non-linear diabetic model incorporating physical, mental, and lifestyle factors.
• To design and implement a novel RECCo controller using an ANYA fuzzy rule-based system for adaptive blood glucose regulation in Type 1 DM patients.
• To validate the controller's performance through hardware experiments and comparative analysis.

Main Methods:

• Development of a Sorensen-based diabetic model considering physical characteristics, mental state, stress, meals, exercise, and Insulin Sensitivity (IS).
• Design of a RECCo controller based on the ANYA fuzzy rule-based system for online adaptive control.
• Implementation of a simple insulin pump for hardware experiments.
• Validation of model accuracy using the N-BEATS algorithm (98% accuracy).
• Comparative analysis with Model Predictive Control (MPC) and Model Reference Adaptive Control (MRAC).

Main Results:

• Successful implementation of the RECCo controller in a hardware experiment for effective blood glucose regulation.
• Demonstrated prevention of hypoglycemia and hyperglycemia through precise glucose control.
• The proposed controller showed superior performance compared to MPC and MRAC in comparative analysis.
• The N-BEATS algorithm validated the model's accuracy at approximately 98%.

Conclusions:

• The developed Sorensen-based diabetic model and RECCo controller offer a robust solution for managing Type 1 DM.
• The adaptive nature of the ANYA fuzzy rule-based system enables effective glucose regulation under uncertain conditions.
• Hardware validation confirms the practical applicability and success of the controller in preventing diabetic complications.