Work

Sparse-Sensor Axial Temperature Reconstruction in Catalytic CO₂ Methanation: Reproducible Software and Results

Year: 2026
Type: article
Abstract: This reproducible computational research archive provides Python code, derived numerical results, figures and verification files for axial temperature reconstruction from three virtual measurement stations in catalytic CO₂ methanation reactors. Published experimental profiles of nickel-based catalysts serve as inputs; original measurements are not redistributed. The evaluation uses leave-one-operating-condition-family-out testing and compares shape-based estimators calibrated using sparse observations against piecewise cubic Hermite interpolation (PCHIP). The archive contains calibration partitions, per-profile and per-family errors, comparisons of reconstructed temperature maxima, and sensitivity tests involving virtual measurement noise and sensor unavailability. In the evaluated Ni/SiO₂ campaign, the reference PCHIP method obtains a macro-average hidden-temperature MAE of 0.615 K, while the target-calibrated estimator using 16 calibration profiles obtains 0.412 K. Reconstruction of temperature maxima is not improved: the corresponding errors are 0.276 K and 0.395 K, respectively. The study is a retrospective virtual-sensor evaluation, not an operational digital twin or an independently validated industrial safety system. Scripts and instructions reproduce the analysis from the original published datasets (DOIs: 10.5281/zenodo.17358086 and 10.5281/zenodo.17455028).
Language: English
Subfield: Catalysis
Open Access status: green
OpenAlex ID: W7235155680
FWCI:
Cites: 2
Cited by:
Related to: 10