Nonlinear EPSAC predictive control of blood glucose level in diabetic patients for prandial disturbance rejection

The nonlinear EPSAC (NEPSAC) predictive control algorithm has been implemented to regulate blood glucose level in type I diabetic patients, by controlling the insulin infusion rate with a micropump. The advantage of the NEPSAC strategy is that the nonlinear process is never linearized, thus it uses directly the nonlinear patient model. The role of the disturbance filter in model-based predictive control is underlined and its possibility to improve control performance is exploited. A default disturbance filter, consisting of an integrator to ensure zero steady-state error, is compared to an intelligent filter design with output constraints. Controller performance was assessed in terms of its ability to reject typical prandial (meal) disturbances of 20 g, 40 g and 30 g glucose.