研究目的
To develop a comprehensive model of FBG spectral distortions and evaluate the performance of demodulation algorithms under various distortions.
研究成果
The proposed comprehensive model effectively characterizes FBG spectral distortions. Evaluation of eight demodulation algorithms shows that no single algorithm performs best for all distortion types, but Radial-Basis Function, Gaussian Fitting, and Cross-Correlation have relatively better overall performance. The study provides guidance for selecting appropriate algorithms based on distortion characteristics and suggests that the model can aid in developing improved algorithms, particularly for neural network-based approaches.
研究不足
The model does not fully address multi-peaks distortion due to the lack of a theoretical definition for Bragg wavelength in such cases. The study relies on simulations, which may not capture all real-world complexities. The algorithms are evaluated individually for each distortion type, but practical scenarios often involve multiple simultaneous distortions.
1：Experimental Design and Method Selection:
The study uses numerical simulations to evaluate FBG demodulation algorithms. A comprehensive model of spectral distortions is developed, incorporating asymmetrical distortion, multi-peaks distortion, interferometric noise, background noise, and random noise. The model is based on theoretical equations derived from FBG sensing principles.
2：Sample Selection and Data Sources:
Simulated FBG spectra are generated using the proposed model. Parameters such as wavelength shift, distortion types, and noise levels are varied to create diverse datasets for evaluation.
3：List of Experimental Equipment and Materials:
No specific physical equipment is used; the work is simulation-based. The model involves parameters like wavelength, reflectivity, noise levels, and distances in optical fibers.
4：Experimental Procedures and Operational Workflow:
For each type of distortion, simulated spectra are generated. Eight demodulation algorithms (Peak Detection, Bandwidth Middle-Point, Centroid, Gaussian Fitting, Cross-Correlation, Steger Image, Fast Phase Correlation, Radial-Basis Function) are applied to these spectra. Demodulation errors are calculated by comparing estimated Bragg wavelength shifts to known values.
5：Data Analysis Methods:
Demodulation errors are quantified as absolute differences between estimated and actual wavelength shifts. Statistical analysis includes calculating maximum errors, standard deviations, and normalized average errors. Results are visualized through plots and comparisons.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
