Abstract: This work presents a fuzzy logic–based closed-loop system for automated blood glucose regulation using continuous glucose monitoring (CGM) and intelligent insulin delivery. Unlike conventional PID controllers that rely on fixed gains and linear assumptions, the proposed controller uses linguistic rules and adaptive membership functions to relate glucose tracking error and its rate of change to appropriate insulin infusion rates. A validated glucose–insulin dynamic model is used to simulate patient response under realistic meal disturbances, and performance is assessed using metrics such as RMSE, MARD, and Time in Range. Results show that the fuzzy controller significantly reduces postprandial glucose excursions, improves average glycemic control, and lowers the frequency of hypoglycemic events compared with both open-loop operation and PID control. Sensitivity analysis further confirms the robustness of the fuzzy architecture to variations in patient parameters and measurement noise. These findings suggest that fuzzy inference offers a promising alternative to fully model-based control strategies for artificial pancreas applications, especially when physiological variability and uncertainty make precise mathematical modeling difficult.

Keywords: Fuzzy logic controller, Blood glucose regulation, Artificial pancreas, Insulin infusion control, Continuous glucose monitoring, Glucose–insulin model, Glycemic variability, Intelligent control

Downloads: | DOI: 10.17148/IARJSET.2025.1211032