ﮐﻨﺘﺮل ﻗﻨﺪ ﺧﻮن ﺑﯿﻤﺎران دﯾﺎﺑﺘﯽ ﻧﻮع ﯾﮏ: دﯾﺪﮔﺎه ﺗﻄﺒﯿﻘﯽ ﻓﺎزي ﻣﻘﺎوم

Blood Glucose Control for Type 1 Diabetic Patients: Robust Fuzzy Adaptive Approach

ﮐﻨﺘﺮل ﻗﻨﺪ ﺧﻮن ﺑﯿﻤﺎران دﯾﺎﺑﺘﯽ ﻧﻮع ﯾﮏ در ﺣﻀﻮر ﻧﺎﻣﻌﯿﻨﯽﻫﺎي ﺳﺎﺧﺘﺎري و ﺑﯽﺳﺎﺧﺘﺎر در اﯾﻦ ﻣﻘﺎﻟﻪ ﻣﻄﺎﻟﻌﻪ ﺷﺪه اﺳﺖ. در راﺳﺘﺎي اﻓﺰاﯾﺶ ﮐﺎرآﻣﺪي روﯾﮑﺮد ﮐﻨﺘﺮﻟﯽ اراﺋﻪ ﺷﺪه، ﻓﺮض ﺷﺪه ﮐﻪ ﺗﻤﺎم دﯾﻨﺎﻣﯿﮏﻫﺎي ﺗﻮﺻﯿﻒﮐﻨﻨﺪه ﺗﻨﻈﯿﻢ ﻗﻨﺪ ﺧﻮن ﺑﯿﻤﺎران دﯾﺎﺑﺘﯽ ﻧﻮع ﯾﮏ ﮐﺎﻣﻼً ﻧﺎﻣﻌﯿﻦ ﻫﺴﺘﻨﺪ. ﺑﺮاﺳﺎس ﺗﻮاﺑﻊ ﺗﻘﺮﯾﺐزن ﻓﺎزي ﻣﺠﻬﺰ ﺷﺪه ﺑﻪ اﻟﮕﻮرﯾﺘﻢ ﺗﻄﺒﯿﻖ و ﻫﻢﭼﻨﯿﻦ اﺳﺘﻔﺎده از روﯾﮑﺮد ﮐﺎﻫﺶ ﺗﻌﺪاد ﭘﺎراﻣﺘﺮﻫﺎي ﺗﻄﺒﯿﻖ، دﯾﻨﺎﻣﯿﮏﻫﺎي ﻧﺎﻣﻌﯿﻦ ﻣﺪل ﺑﺮﮔﻤﻦ ﺗﻘﺮﯾﺐ زده ﺷﺪه اﺳﺖ. ﺳﭙﺲ ﺑﺮاﺳﺎس روﯾﮑﺮد ﮐﻨﺘﺮل ﺧﻄﯽﺳﺎز ﻓﯿﺪﺑﮏ و ﺟﺒﺮانﺳﺎز ﺗﻄﺒﯿﻘﯽ ﻣﻘﺎوم، ﻃﺮاﺣﯽ ﮐﻨﺘﺮلﮐﻨﻨﺪه ﺧﻄﯽﺳﺎز ﻓﯿﺪﺑﮏ ﻣﻘﺎوم ﻓﺎزي ﺑﻪﻣﻨﻈﻮر ﺗﻨﻈﯿﻢ ﻗﻨﺪ ﺧﻮن ﺑﯿﻤﺎران دﯾﺎﺑﺘﯽ ﻧﻮع ﯾﮏ در ﺣﻀﻮر وﻋﺪه ﻏﺬاﯾﯽ ﺑﺮاي اوﻟﯿﻦ ﺑﺎر ﻣﻮرد ﺑﺮرﺳﯽ ﻗﺮار ﮔﺮﻓﺘﻪ اﺳﺖ. ﺑﺎ اﺳﺘﻔﺎده از ﺗﺌﻮري ﻟﯿﺎﭘﺎﻧﻮﻓﯽ ﻧﺸﺎن داده ﺷﺪه ﮐﻪ ﺗﻤﺎم ﺳﯿﮕﻨﺎلﻫﺎي ﺳﯿﺴﺘﻢ ﺣﻠﻘﻪ ﺑﺴﺘﻪ ﮐﻨﺘﺮل ﻗﻨﺪ ﺧﻮن ﺑﻪﺻﻮرت ﮐﺮاندار ﻧﻬﺎﯾﯽ ﯾﮑﻨﻮاﺧﺖ، ﭘﺎﯾﺪار ﻫﺴﺘﻨﺪ و ﻫﻤﭽﻨﯿﻦ ﻗﻨﺪ ﺧﻮن ﺑﯿﻤﺎران دﯾﺎﺑﺘﯽ ﺑﻪ ﮐﺮاﻧﯽ از ﻣﻘﺪار ﻣﻄﻠﻮب ﻫﻤﮕﺮا ﺷﺪه اﺳﺖ. در اﻧﺘﻬﺎ، ﻧﺘﺎﯾﺞ ﺷﺒﯿﻪﺳﺎزي ﻧﺸﺎندﻫﻨﺪه ﻋﻤﻠﮑﺮد ﻣﻄﻠﻮب ﮐﻨﺘﺮلﮐﻨﻨﺪه در ﮐﺎﻫﺶ اﺛﺮ اﻏﺘﺸﺎش وﻋﺪه ﻏﺬاﯾﯽ و ﻫﻤﭽﻨﯿﻦ رﻓﺘﺎر ﻣﻘﺎوم در ﺑﺮاﺑﺮ دﯾﻨﺎﻣﯿﮏﻫﺎي ﻧﺎﻣﻌﯿﻦ و ﺧﻄﺎي ﺗﺨﻤﯿﻦ وﻋﺪه ﻏﺬاﯾﯽ اﺳﺖ. ﻫﻢﭼﻨﯿﻦ، ﻋﻤﻠﮑﺮد ﻣﻄﻠﻮب ﮐﻨﺘﺮلﮐﻨﻨﺪه ﭘﯿﺸﻨﻬﺎد ﺷﺪه در ﮐﻨﺘﺮل ﻗﻨﺪ ﺧﻮن )ﻧﮕﻪ داﺷﺘﻦ ﻗﻨﺪ ﺧﻮن در ﻣﺤﺪوده ﻣﺠﺎز 70-120 ﻣﯿﻠﯽ-ﮔﺮم در دﺳﯽ ﻟﯿﺘﺮ( در ﻣﻘﺎﯾﺴﻪ ﺑﺎ ﺑﺮﺧﯽ از ﻣﺮاﺟﻊ ﻧﺸﺎن داده ﺷﺪه اﺳﺖ.

In this paper, Blood Glucose control in type 1 diabetic patients in the presence of structured and unstructured uncertainties is studied. In order to increase the effectiveness of the proposed control approach, assumed that all the dynamics describing the regulation of Blood Glucose in type 1 diabetic patients are completely unknown. Based on the fuzzy approximation function, which is equipped with the adaptive algorithm and employing the approach of reducing the number of adaptive fuzzy parameters, the unknown dynamics of the Bergman model approximated. Then, based on the feedback linearization control approach and robust adaptive compensator, the design of feedback linearization robust fuzzy controller to regulate Blood Glucose in type 1 diabetic patients in the presence of meal is studied for the first time. Using Lyapunov theory, it is shown that all signals of the closed-loop system are uniformly ultimately bounded and the Blood Glucose of diabetic patients converges to the neighborhood of the desired value. Finally, the simulation results show a good controller performance in reducing effect of the meal disturbance, and robustness against uncertain dynamics and meal estimation errors. Moreover, in comparison with some existing results, a good performance of the introduced controller in controlling Blood Glucose of diabetic patients (i.e., keeping Blood Glucose in allowed range 70-120 mg/dl) validated.