Intelligent insulin injection system for the diabetic patients based on neural network and fuzzy algorithm

Prcision and stable control of blood glucose level is an essential requirement for Diabetic patients. On the other hand, biological systems (e.g. Insulin-glucose interaction) act as a non-deterministic and nonlinear complex model. Therefore, we need to use intelligent control. In this paper we use a practical model (instead of the patient) and simulate a neural-fuzzy controller for blood glucose adjustment. Neural network estimates the future of the model then generates the control signal (insulin injection rate). This is a Model Prediction Control (MPC). Fuzzy controller acts as a supervisor system and guarantees the stable and robust control.