| Autor: |
Djeziri, Mohand, Benmoussa, Samir, Occelli, Clement, Fiorido, Tomas, Seguin, Jean-Luc, Bendahan, Marc |
| Zdroj: |
IFAC-PapersOnLine; January 2024, Vol. 58 Issue: 4 p479-483, 5p |
| Abstrakt: |
Power-to-Gas (P2G) technology converts renewable energy into hydrogen through electrolysis and integrates it into the natural gas network. However, controlling hydrogen concentrations in the hydrogen/natural gas mixture is challenging for storage, transportation, and utilisation. Machines and turbines designed for natural gas operation have specific hydrogen concentration limits and cannot accommodate significant variations in the mixture. Therefore, forecasting the time evolution of the hydrogen concentration in the natural gas network with a sufficiently long forecasting horizon before supplying the machines is essential. This paper proposes a data-driven solution that combines a novel hydrogen sensor technology with geometric prediction models to forecast hydrogen concentration. The sensor uses a sensitive layer of nanostructured palladium-gold (PdAu) alloy deposited via cathode sputtering. The sensor’s raw measurements serve as inputs to the prediction models, which rely on online position and velocity calculations. These geometric models are particularly suitable for gas concentration forecasting as they are continuously updated based on the online position and velocity calculations and do not require prior knowledge of the concentration profile. Experimental results from a test bench demonstrate the effectiveness of the proposed sensors and the significance of the prediction method. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
