Robust LPV control design for blood glucose regulation considering daily life factors. (March 2020)

• Record Type: Journal Article
• Authors: Mirzaee, Alireza; Dehghani, Maryam; Mohammadi, Mohsen
• Is Part Of: Biomedical signal processing and control. Volume 57(2020)
• Publication Date: 2020-03
• Subjects: Diabetes -- Glucose-insulin model -- Life factors -- Linear parameter varying (LPV) control -- Stochastic simulation; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2019.101830
• Languages: English
• ISSNs: 1746-8094

Abstract

Graphical abstract: Highlights: A Linear Parameter Varying (LPV) model of the glucose-insulin system is introduced. Effects of food intake, exercise, stress and fatigue on type 1 diabetic patients are investigated, simultaneously. A robust linear parameter varying control approach considering the primary and secondary everyday life factors is proposed. Postprandial CVGA chart is introduced which uses a bond-reject filter around the maximum values in the blood glucose curve. The introduced PCVGA, is more intuitive and less conservative than CVGA for the extreme glucose deviations classification. Abstract: In blood glucose control, the daily life factors such as food intake, exercise, stress, fatigue and circadian rhythm that alter the level of blood glucose should be meticulously considered, which requires the design of a robust control approach for dealing with these uncertainties. In this paper, a robust LPV-based controller is designed in order to achieve a predefined level of blood glucose considering the effects of daily life factors. To be sure that the proposed control signal, which shows the insulin dosage, is practical, a filter with optimized parameters is added to limit the lower and upper bounds of the required exogenous insulin. The suggested approach is applied on an in-silico simulated model of a diabetic patient and the results are presented which show the efficacy of the algorithm for insulin dosage determination. The effects of several everyday life scenarios including announced and random unannounced meals and snacks, stress, fatigue, exercise and circadian rhythm are studied through various simulations and the results are presented.