Bayesian Optimization-Based Tunable Explicit MPC on a Pocket-Sized Embedded Platform
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F25%3A00387693" target="_blank" >RIV/68407700:21230/25:00387693 - isvavai.cz</a>
Výsledek na webu
<a href="http://dx.doi.org/10.1109/CDC57313.2025.11312890" target="_blank" >http://dx.doi.org/10.1109/CDC57313.2025.11312890</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1109/CDC57313.2025.11312890" target="_blank" >10.1109/CDC57313.2025.11312890</a>
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Bayesian Optimization-Based Tunable Explicit MPC on a Pocket-Sized Embedded Platform
Popis výsledku v původním jazyce
The paper presents a pocket-sized embedded platform designed for the validation of advanced control methods. The platform is based on the ESP32-S3 microcontroller and is equipped with a miniaturized heat exchange device, making it suitable for temperature control experiments. The platform enables the implementation of a real-time tunable explicit Model Predictive Control (MPC) algorithm, which allows for online tuning of the weighting parameter in the MPC cost function. The paper also introduces a novel application of a method based on Bayesian optimization, which efficiently explores the parameter space to find an optimal performance metric value. The performance metric is a weighted overall parameter of the controller’s performance, in which actuator power consumption and control signal fluctuation are evaluated. Experimental results demonstrate the effectiveness of the proposed approach, while the exploration and exploitation approaches have been tested. Both the explicit MPC controller and the Bayesian optimization method are implemented on the embedded platform, showcasing its capabilities for real-time control applications. The results highlight the potential of the platform for further research and development in advanced control strategies.
Název v anglickém jazyce
Bayesian Optimization-Based Tunable Explicit MPC on a Pocket-Sized Embedded Platform
Popis výsledku anglicky
The paper presents a pocket-sized embedded platform designed for the validation of advanced control methods. The platform is based on the ESP32-S3 microcontroller and is equipped with a miniaturized heat exchange device, making it suitable for temperature control experiments. The platform enables the implementation of a real-time tunable explicit Model Predictive Control (MPC) algorithm, which allows for online tuning of the weighting parameter in the MPC cost function. The paper also introduces a novel application of a method based on Bayesian optimization, which efficiently explores the parameter space to find an optimal performance metric value. The performance metric is a weighted overall parameter of the controller’s performance, in which actuator power consumption and control signal fluctuation are evaluated. Experimental results demonstrate the effectiveness of the proposed approach, while the exploration and exploitation approaches have been tested. Both the explicit MPC controller and the Bayesian optimization method are implemented on the embedded platform, showcasing its capabilities for real-time control applications. The results highlight the potential of the platform for further research and development in advanced control strategies.
Klasifikace
Druh
D - Stať ve sborníku
CEP obor
—
OECD FORD obor
20205 - Automation and control systems
Návaznosti výsledku
Projekt
<a href="/cs/project/EH22_008%2F0004590" target="_blank" >EH22_008/0004590: Robotika a pokročilá průmyslová výroba</a><br>
Návaznosti
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Ostatní
Rok uplatnění
2025
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Údaje specifické pro druh výsledku
Název statě ve sborníku
Proceedings of the IEEE 64th Conference on Decision and Control
ISBN
979-8-3315-2627-6
ISSN
0743-1546
e-ISSN
2576-2370
Počet stran výsledku
8
Strana od-do
4663-4670
Název nakladatele
IEEE
Místo vydání
Piscataway
Místo konání akce
Rio de Janeiro
Datum konání akce
10. 12. 2025
Typ akce podle státní příslušnosti
WRD - Celosvětová akce
Kód UT WoS článku
—