Model bias and calibration uncertainty in dispersive interrogation of fiber-optic Fabry–Perot sensors
Measurement Science and Technology · 16.11.2026
DOI: 10.1088/1361-6501/aec28a ↗Kurz gefasst
Fachartikel in der Zeitschrift Measurement Science and Technology, online erschienen am 16.11.2026. Verfasst von Hang Zheng und 6 weiteren Autor:innen. Gefördert von: the Special Research Fund of the National Key Laboratory of Intense Pulsed Radiation Simulation and Effect, the Hainan Province Science and Technology Special Fund.
Abstract (Originalsprache)
Dispersive interrogation converts Fabry–Perot spectra into temporal waveforms, but cavity-length estimates depend on both the propagation approximation and the calibration of the time–wavelength coordinate. We formulate a coherent two-beam forward model retaining relative delay, reflection phase, second- and third-order dispersion, and detector bandwidth. Variable projection removes pulse-dependent gain and offset, yielding a local expression that relates waveform mismatch to length bias and propagates correlated calibration uncertainties. An information analysis quantifies the contribution of an independent phase observation. In a controlled 50 μm cavity simulation at a signal-to-noise ratio of 60 dB, 200 trials give root-mean-square length errors of 7.980, 0.0279 and 0.0060 nm for linear intensity mapping, nonlinear intensity mapping and field-template estimation, respectively, with calibration parameters fixed at their true values. The corresponding local standard-deviation bound increases from 0.0060 to 2.548 nm when reflection phase is unknown. Separate measurements test coordinate calibration against an optical spectrum analyser. For an unloaded stainless-steel specimen, a dispersion-constrained local linear mapping reduces the normalized spectral residual from 7.44% to 3.93% and the validation-peak wavelength discrepancy from 0.317 to 0.053 nm. For a sapphire specimen, physical nonlinear mapping reduces spectral residuals from 6.32–6.63% to 4.94–5.18% in two static validation windows; empirical quadratic calibration gives 4.67–4.94%. The results separate errors due to reference-wavelength scaling, mapping curvature and finite-dispersion propagation, and quantify the phase calibration needed to benefit from a coherent estimator.
Förderung
- the Special Research Fund of the National Key Laboratory of Intense Pulsed Radiation Simulation and Effect
- the Hainan Province Science and Technology Special Fund
Zitieren
APA
Zheng, H., Zhu, Z., Wang, Z., Liu, W., Yang, J., Chen, B., & Zhang, D. (2026). Model bias and calibration uncertainty in dispersive interrogation of fiber-optic Fabry–Perot sensors. Measurement Science and Technology. https://doi.org/10.1088/1361-6501/aec28a
Vancouver
Zheng H, Zhu Z, Wang Z, Liu W, Yang J, Chen B, et al. Model bias and calibration uncertainty in dispersive interrogation of fiber-optic Fabry–Perot sensors. Measurement Science and Technology. 2026. doi:10.1088/1361-6501/aec28a
BibTeX
@article{Zheng2026,
author = {Zheng, Hang and Zhu, Zhaolin and Wang, Zhenyu and Liu, Wenxiang and Yang, Jun and Chen, Bin and Zhang, Dezhi},
title = {Model bias and calibration uncertainty in dispersive interrogation of fiber-optic Fabry–Perot sensors},
journal = {Measurement Science and Technology},
year = {2026},
doi = {10.1088/1361-6501/aec28a}
}
EUVIMED