This paper proposes an alternative approach to the signal processing of temperature measurements based on Fiber Bragg Gratings (FBG) using the machine learning tool Gaussian Process Regression (GPR). Experimental results show that for a majority of the cases under consideration the reported technique provides a more accurate calculation of the temperature than the conventional methods. Furthermore, the GPR can give the uncertainty of an estimate together with the estimate itself, which for example is useful when it is important to know the worst case scenario of a measurand. GPR also has the potential to improve the measurement speed of FBG based temperature sensing compared to current standards.
16 September 2022 in Optics and Lasers in Engineering
The authors report on an analytic connector loss model for step-index polymer optical fibers (SI-POF). The model describes the non-linear dependencies for combinations of the most important intrinsic and extrinsic influence parameters like Fresnel reflections, mismatches in numerical aperture (NA) and core diameters, as well as lateral, longitudinal and angular offsets. As connector losses strongly depend on the mode distribution, expressions for the mode-dependent coupling efficiencies for the various influence parameters are derived. This way, the model can handle any rotationally symmetric mode distribution including the uniform mode distribution (UMD) and the equilibrium mode distribution (EMD) in the steady-state. It is shown that for step-index fibers, lateral and longitudinal offsets can be handled in a similar way, which reduces the overall effort for the mathematical description. For the typical case of identical fibers, the authors also derive easy-to-apply approximations for a combination of lateral and longitudinal offsets under UMD and EMD conditions. The results are in good agreement with ray-tracing simulations and are evaluated for the cases of identical fibers and the worst case parameter constellation for the SI-POF in the IEC A4a.2 fiber class.
The intrinsic and extrinsic coupling losses of step-index polymer optical fibers are statistically examined by Monte Carlo simulations. In contrast to most existing models that linearly scale individual losses, a comprehensive analytic coupling loss model is used that also considers the interdependencies between mismatches in numerical aperture and core diameter, as well as radial and longitudinal offsets. As a typical example, the coupling losses of A4a.2 step-index multimode fibers are analyzed for an equilibrium mode distribution. The results show considerably less conservative coupling loss estimations than with traditional models, improving link power budgeting.
We propose an efficient model-based signal processing approach for optical fiber sensing with fiber Bragg grating (FBG) arrays. A position estimation based on an estimation of distribution algorithm (EDA) and a reflectivity estimation method using a parametric transfer matrix model (TMM) are outlined in detail. The estimation algorithms are evaluated with Monte Carlo simulations and measurement data from an incoherent optical frequency domain reflectometer (iOFDR). The model-based approach outperforms conventional Fourier transform processing, especially near the spatial resolution limit, saving electrical bandwidth and measurement time. The models provide great flexibility and can be easily expanded in complexity to meet different topologies and to include prior knowledge of the sensors. Systematic errors due to crosstalk between gratings caused by multiple reflections and spectral shadowing could be further considered with the TMM to improve the performance of large-scale FBG array sensor systems.
A method for time and wavelength division multiplexing of fiber Bragg gratings using an electro-optical modulator in a bidirectional configuration is proposed. Broadband frequency conversion enables position measurement while an optical spectrum analyzer provides wavelength resolution.