A Spectral Analysis Method Based on Direct Absorption Intensity Reconstruction for Rapid Real-Time CO Detection

During combustion processes, carbon monoxide concentrations typically undergo rapid fluctuations within seconds. Consequently, real-time detection of trace CO concentrations in combustion scenarios holds significant importance for combustion safety alerts and human health. Given its advantages in measurement timeliness and system stability, direct absorption spectroscopy (DAS) technology is considered an ideal approach for achieving rapid, real-time CO detection in combustion environments. However, DAS exhibits slow fitting processes and extreme sensitivity to various optical noises, limiting the accuracy of concentration inversion. Existing algorithms lack standardized usage protocols. To address these challenges, this study introduces a spectral analysis method based on direct absorption intensity reconstruction for rapid real-time CO detection. This method does not need to introduce complex modulation techniques and large-scale numerical operations, and can be achieved only based on the original measurement signal. Simultaneously, usage standards for reference signals were established by calculating the root-mean-square error (RMSE) and correlation coefficient (ρ). Experimental results indicate that when the signal-to-noise ratio (SNR) of the reference signal used for reconstruction exceeds 32, further increases in SNR have negligible impact on the reconstructed signal. The RMSE between reconstructed and measured results is consistently below 1 × 10^-3 with ρ exceeding 0.95. The R² value for concentration inversion relative to absorbance rose to 0.9998. SNR improved by over 2-fold. In real-time monitoring, the measurement system delivers rapid responses within 4 s, meeting the requirements for fast, real-time CO detection in combustion scenarios.