CH₄, C₂H₆, and CO₂ Multi-Gas Sensing Based on Portable Mid-Infrared Spectroscopy and PCA-BP Algorithm

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Sensor Technology

Background:

• Accurate detection of multiple gases like methane (CH4), ethane (C2H6), and carbon dioxide (CO2) is crucial for environmental monitoring and industrial safety.
• Existing methods may face challenges with cross-interference and cost-effectiveness for multi-component gas analysis.

Purpose of the Study:

• To develop and validate a multi-gas sensing system capable of simultaneously detecting CH4, C2H6, and CO2.
• To address the challenge of spectral interference between CH4 and C2H6 using advanced algorithms.

Main Methods:

• Utilized a non-dispersive infrared (NDIR) detection principle.
• Employed a broad-spectra light source, a tunable Fabry-Pérot (FP) filter detector, and an infrared waveguide as an absorption cell.
• Implemented a PCA-BP neural network algorithm to resolve interference and quantify CH4 and C2H6 concentrations.

Main Results:

• Achieved detection limits of 2.59 ppm for CH4, 926 ppb for C2H6, and 114 ppb for CO2.
• Reported Root Mean Square Errors of Prediction (RMSEP) of 10.97 ppm for CH4 and 2.00 ppm for C2H6.
• Demonstrated the system's ability to leverage a mid-infrared broadband source for multi-component gas measurement.

Conclusions:

• The developed NDIR-based system effectively detects multiple gases with high sensitivity and accuracy.
• The integration of a PCA-BP neural network algorithm successfully mitigates spectral interference.
• The system presents a balanced approach between performance and cost for multi-gas sensing applications.