Demodulation algorithm for temperature and pressure sensing in side-hole fiber Bragg gratings using a CNN-LSTM...

Area of Science:

• Photonics and Optical Sensing
• Artificial Intelligence in Engineering
• Materials Science for Extreme Environments

Background:

• Side-hole fiber Bragg gratings (SHFBG) offer simultaneous temperature and pressure sensing in confined spaces.
• Combined temperature and pressure effects complicate SHFBG spectral analysis, challenging traditional demodulation algorithms.
• Accurate simultaneous measurement is critical for industries like petroleum and aviation.

Purpose of the Study:

• To develop a high-precision deep learning model for simultaneous temperature and pressure demodulation from SHFBG.
• To address the spectral overlap and separation issues caused by combined environmental factors.
• To enhance the accuracy and reliability of SHFBG sensing in challenging industrial settings.

Main Methods:

• Implementation of a Convolutional Neural Network (CNN) combined with a Long Short-Term Memory (LSTM) recurrent neural network architecture.
• Training and validation of the CNN-LSTM model on SHFBG spectral data, including datasets with low wavelength resolution and high signal-to-noise ratio.
• Comparative analysis against various neural network architectures to evaluate performance.

Main Results:

• The CNN-LSTM model achieved a root mean square error (RMSE) of 0.296 MPa for pressure and 0.276°C for temperature on simulated data.
• Experimental measurements showed errors of 2.425°C for temperature and 2.243 MPa for pressure.
• The proposed model demonstrated superior training accuracy and generalization capability compared to other neural network models.

Conclusions:

• The CNN-LSTM deep learning model effectively overcomes the limitations of traditional algorithms for SHFBG demodulation.
• This approach enables high-precision simultaneous measurement of temperature and pressure in confined environments.
• The research facilitates broader applications and further development of SHFBG sensing technologies.