Constrained Versus Unconstrained Model Predictive Control for Artificial Pancreas

The technological advances reached in the last years released portable devices with high computational capabilities and able to overcome relevant hardware limitations of past artificial pancreas (AP) applications. In view this, choice an unconstrained saturated model predictive control (S-MPC) can be reconsidered. A constrained MPC (C-MPC) is formulated here as a finite-horizon optimal problem it retuned, well S-MPC, using new UVA/Padova simulator. calibration procedure used consider clinically significant performance indices. C-MPC S-MPC are tested silico on 100 adult patients Its capability better represent real-life conditions allows evaluate MPCs more realistic scenarios. outperforms terms average glucose, time spent tight range, above 180 mg/dL. An acceptable increase below 70 mg/dL 4% present for 25% patients, but severe hypoglycemia remains equal 0%. development patient-tailored proposed future mitigate these phenomena.