Related Experiment Video
Updated: Jul 15, 2025
00:08
A Random-displacement Measurement by Combining a Magnetic Scale and Two Fiber Bragg Gratings
Published on: September 30, 2019
6.3K
Distributed curvature sensing using long period fiber grating and machine learning numerical analysis.
Optics Letters
|September 29, 2023
Summary
This study introduces a novel fiber optic sensor using a long period fiber grating (LPG) and neural networks to accurately detect fiber bending. The system achieves a low 0.40% error for curvature profile estimation.
Area of Science:
- Photonics and Optical Sensing
- Artificial Intelligence in Engineering
- Fiber Optic Sensors
Background:
- Long Period Fiber Gratings (LPGs) are sensitive to external perturbations.
- Accurate measurement of fiber curvature is crucial for various applications.
- Traditional methods for curvature sensing can be complex and less precise.
Purpose of the Study:
- To propose and numerically validate a fiber distributed curvature sensor.
- To leverage neural networks for analyzing LPG spectral transmission data.
- To assess the sensor's performance in detecting non-uniform bending profiles.
Main Methods:
- Numerical simulation of a 6-cm LPG structure using EigenMode Expansion and coupled-mode theory.
- Analysis of optical transmission spectra corresponding to different fiber curvature profiles.
- Training a four-dense-layer neural network with simulated data.
Main Results:
- The developed neural network model achieved a 0.40% relative median estimation error for bending profiles.
- Demonstrated the capability to analyze non-uniform perturbations effectively.
- Validated the potential of LPGs in conjunction with AI for sensing applications.
Conclusions:
- Neural network-based optical sensors are efficient for analyzing non-uniform perturbations.
- Long-period fiber gratings are promising candidates for advanced curvature sensing systems.
- This approach offers a high-accuracy, AI-driven solution for fiber optic sensing.
