Design of PSO Based Robust Blood Glucose Control in Diabetic Patients

Abstract

Abstract – In this paper, the design of robust blood glucose controller in diabetes using H-infinity technique is presented. The Particle Swarm Optimization (PSO) method is used to tune the specific structure controller parameters subject to H-infinity constraints. The Bergman model is used to represent the artificial pancreas. This model is one of the more widely used models of the effect of insulin infusion and glucose inputs on the blood glucose concentration. The results show the effectiveness of the designed controller in controlling the behavior of glucose deviation to a sudden rise in the blood glucose. The proposed controller can effectively attenuate the blood glucose deviation to 0.15. This value of attenuation makes the proposed controller superior to the other controllers in previous works. Matlab 7.11 is used to demonstrate the simulation results.