An intelligent approach for optimal regulation of blood glucose level

Diabetes mellitus, a situation resulting from the inability of the pancreas in controlling blood glucose level, can lead to serious long-term complications and death. In this paper, the model of type-1 diabetes mellitus has been studied considering the inputs of the system which are daily meals and the amount of insulin injected. With the help of particle swarm optimization algorithm and the use of proposed controller based on the Hammerstein model, an optimized value of injected insulin has been determined in order to control blood glucose level. Results obtained have been compared to results of the classic controller, and that demonstrates improvement in operation of the proposed controller´s algorithm.