- Titel
- A Model-Based Approach for Online Optimization of Pneumatic Drives
- Zitierfähige Url:
- https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-964399
- Quellenangabe
- IEEE ASME transactions on mechatronics
Erscheinungsjahr: 2024
Seiten: 1-11
E-ISSN: 1941-014X
Artikelnummer: 104 - Erstveröffentlichung
- 2024
- Abstract (EN)
- This article presents an online monitoring and setup strategy to optimize the operating conditions of pneumatic drives. The strategy integrates a model-based optimization approach with a hybrid machine learning (HML) model designed to assess the loading conditions and improve the performance of pneumatic actuators. Employing the operating point method, a model-based approach, the HML model seeks an optimal balance between energy efficiency and robustness. A strategy based on the chambers’ pressures is applied on the HML model to assess in real-time the actual load being applied on the cylinder, establishing a reference pair of pressure ratios, and making it suitable for nonconstant load force applications. A sensitivity analysis is performed to evaluate the impact of the uncertainties associated with the operating point parameters on the HML model output. Experimental results under several load conditions demonstrated its ability to significantly reduce air consumption in underloaded scenarios. Furthermore, the HML approach improved system robustness in overloaded cases, mitigating the impact on displacement time due to possible load variations. Comparative evaluations with a conventional architecture and a commercially available energy-saving device highlight the capabilities of the proposed online monitoring strategy of cost-effectively enhancing the operation of pneumatic drives.
- Andere Ausgabe
- Link zum Artikel, der zuerst in der Zeitschrift „ IEEE ASME transactions on mechatronics” bei IEEE erschienen ist.
DOI: 10.1109/TMECH.2024.3510220 - Verweis
- Ergänzendes Material ist unter folgendem Link zu finden.
DOI: 10.1109/TMECH.2024.3510220. - Freie Schlagwörter (EN)
- Energy efficiency, machine learning, online monitoring, operating point, pneumatic systems, robustness
- Klassifikation (DDC)
- 620
- Verlag
- IEEE, New York, NY
- Version / Begutachtungsstatus
- angenommene Version / Postprint / Autorenversion
- URN Qucosa
- urn:nbn:de:bsz:14-qucosa2-964399
- Veröffentlichungsdatum Qucosa
- 03.04.2025
- Dokumenttyp
- Artikel
- Sprache des Dokumentes
- Englisch
- Lizenz / Rechtehinweis
