PSO-MPC control of artificial pancreas

Abstract

PSO tuning of Model Predictive Control (MPC) implemented in Artificial Pancreas is presented in this paper. The objective function of PSO-based MPC is defined using closed loop information to perform a minimal computation time which is achieved by a small number of iterations and a reduced population size. The controller implementation in a test platform approved by the FDA shows an acceleration in reference tracking with a reduction of overshoot and a complete avoidance of hypoglycemia.