Glycemic Control and Temperature Control in Buildings
Author
Summary, in English
This thesis consists of two parts, applying concepts from automatic control to different application areas.
The first part of this thesis concerns the development of an optimization-based algorithm, determining the size of insulin and glucose doses for patients suffering from Diabetes Mellitus and treated with multiple insulin injections. Diabetes Mellitus is a chronic disease characterized by elevated blood glucose levels. The therapy usually consists of insulin injections, where the amount of insulin to be administered is decided by the patient using empirically developed rules of thumb. An algorithm is proposed determining the dose intakes of insulin and glucose bringing the blood glucose concentration back to a healthy range. The algorithm uses optimization methods and patient-individual blood glucose predictions to determine these
doses. The cost function used for the optimization problem reflects the risk associated with the blood glucose values. A virtual patient was used as an in-silico test-bed for the proposed algorithm. The results were compared to a bolus calculator. It was found that the proposed control algorithm could improve the time of the simulated patient’s blood glucose spends in a safe range compared to the bolus calculator.
The second part of this thesis aims at applying inverted decoupling to the area of temperature control in buildings. With inverted decoupling, a multi-variable system can be controlled as several single input single output systems. Buildings are multi-variable systems with many interacting variables. In the second part of this thesis, inverted decoupling is applied to two examples to decrease couplings in the dynamics. For
the first example, the aim was to use the decoupling method to decrease interactions of the temperature dynamics of adjacent rooms, in order to be able to regulate the temperature of each room without influence from another room. In the second example, the room temperature was to be regulated using the temperature of the air in a ventilation system. However, changing the air flow rate in the ventilation system influences the room temperature as well. The aim was to use inverted decoupling to decrease this coupling. In simulation studies, the proposed decoupled controller could reduce the effect of the couplings in both examples.
The first part of this thesis concerns the development of an optimization-based algorithm, determining the size of insulin and glucose doses for patients suffering from Diabetes Mellitus and treated with multiple insulin injections. Diabetes Mellitus is a chronic disease characterized by elevated blood glucose levels. The therapy usually consists of insulin injections, where the amount of insulin to be administered is decided by the patient using empirically developed rules of thumb. An algorithm is proposed determining the dose intakes of insulin and glucose bringing the blood glucose concentration back to a healthy range. The algorithm uses optimization methods and patient-individual blood glucose predictions to determine these
doses. The cost function used for the optimization problem reflects the risk associated with the blood glucose values. A virtual patient was used as an in-silico test-bed for the proposed algorithm. The results were compared to a bolus calculator. It was found that the proposed control algorithm could improve the time of the simulated patient’s blood glucose spends in a safe range compared to the bolus calculator.
The second part of this thesis aims at applying inverted decoupling to the area of temperature control in buildings. With inverted decoupling, a multi-variable system can be controlled as several single input single output systems. Buildings are multi-variable systems with many interacting variables. In the second part of this thesis, inverted decoupling is applied to two examples to decrease couplings in the dynamics. For
the first example, the aim was to use the decoupling method to decrease interactions of the temperature dynamics of adjacent rooms, in order to be able to regulate the temperature of each room without influence from another room. In the second example, the room temperature was to be regulated using the temperature of the air in a ventilation system. However, changing the air flow rate in the ventilation system influences the room temperature as well. The aim was to use inverted decoupling to decrease this coupling. In simulation studies, the proposed decoupled controller could reduce the effect of the couplings in both examples.
Department/s
Publishing year
2016
Language
English
Publication/Series
PhD Thesis TFRT-1114
Full text
- Available as PDF - 1 MB
- Available as PDF - 97 kB
- Download statistics
Document type
Dissertation
Publisher
Department of Automatic Control, Lund Institute of Technology, Lund University
Topic
- Control Engineering
- Other Medical Biotechnology
Keywords
- Diabetes, Prediction, Optimization, Decision Support System, Inverted Decoupling, PID, Buildings, Ventilation System
Status
Published
Project
- LCCC
Supervisor
ISBN/ISSN/Other
- ISSN: 0280-5316
- ISBN: 978-91-7623-865-3
- ISBN: 978-91-7623-864-6
Defence date
17 June 2016
Defence time
10:15
Defence place
Lecture hall M:B, builing M, Ole Römers väg 1, Lund University, Faculty of Engineering LTH, Lund
Opponent
- Jan Sternby (Professor)