Application of NARX dynamic neural network in blood glucose prediction model

Establishing a blood glucose prediction model and controlling the amount of insulin injected according to the predicted blood glucose value can effectively reduce the harm caused by high and low blood glucose to the human body. Because the dynamic process of glucose-insulin metabolism in human body is a typical non-steady-state and non-linear process, and has a large time-lag dynamic characteristic, it is difficult for linear models to accurately describe the blood glucose-insulin model. With the development of artificial neural network and its application in all aspects of modern times, this paper proposes to use NARX dynamic neural network to build blood glucose-insulin model, and use the data generated by UVa/Padova simulation platform for network training. Finally, through MATLAB simulation verification, by calculating the mean square error and Clark error grid analysis method, it is shown that the blood glucose can be effectively predicted.

[1]  Liu Ya-qiu,et al.  NARX network for dynamic system identification in adaptive inverse control , 2005 .

[2]  W. A. Sandham,et al.  Neural network and neuro-fuzzy systems for improving diabetes therapy , 1998, Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Vol.20 Biomedical Engineering Towards the Year 2000 and Beyond (Cat. No.98CH36286).

[3]  Youqing Wang,et al.  Hypoglycemia prediction using extreme learning machine (ELM) and regularized ELM , 2013, 2013 25th Chinese Control and Decision Conference (CCDC).

[4]  Wang Youqing,et al.  A novel adaptive run-to-run control strategy for insulin pump therapy optimization , 2013, Proceedings of the 32nd Chinese Control Conference.

[5]  Gregory L. Plett,et al.  Adaptive inverse control of linear and nonlinear systems using dynamic neural networks , 2003, IEEE Trans. Neural Networks.

[6]  Meng Song.  Research on Modeling of Shock Accelerator Based on NARX Neural Network , 2008 .

[7]  Chunhui Zhao,et al.  Blood glucose control based on rapid model identification with particle swarm optimization method , 2017, 2017 29th Chinese Control And Decision Conference (CCDC).

[8]  Giuseppe De Nicolao,et al.  Neural Network Incorporating Meal Information Improves Accuracy of Short-Time Prediction of Glucose Concentration , 2012, IEEE Transactions on Biomedical Engineering.

[9]  Wang Youqing,et al.  Dynamic model with time varying delay for type 1 diabetes mellitus identified by using expectation maximization algorithm , 2016, 2016 35th Chinese Control Conference (CCC).