PID Control Based on BP Neural Network for the Regulation of Blood Glucose Level in Diabetes

Chengwei Li,Ruiqiang Hu
DOI: https://doi.org/10.1109/bibe.2007.4375709
2007-01-01
Abstract:Good control of blood glucose levels significantly reduces vascular complications in diabetes. In this paper, a theoretical analysis of the control of blood glucose levels in diabetic individuals is undertaken using a simple mathematical model of the dynamics of glucose and insulin interaction in the blood system developed by Bergman et al. A PID controller based on BP neural network is used as a better alternative to robustly stabilize the blood glucose level of a diabetic patient. For comparison purpose, a classical PID controller has been designed. The aim of the control performance was assessed in terms of its ability to track a setpoint (70 mg/dL) form initial state of hyperglycemia, as well as in response to a meal disturbance. The MATLAB simulation is performed to manifest the theoretical analysis. The development reported in this paper is expected to simplify the insulin delivery, thereby enhancing the quality of life of the patient.
What problem does this paper attempt to address?