The research aims to address the issue of noise interference affecting the second harmonic signal in quartz-enhanced photoacoustic spectroscopy (QEPAS) gas detection system. We propose a novel signal denoising method that integrates several advanced techniques: improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN), permutation entropy (PE), and wavelet threshold denoising (WTD). The method is implemented as follows: firstly, the ICEEMDAN algorithm is employed to break down the second harmonic signal into multiple intrinsic mode functions (IMFs). Secondly, the PE values for each IMF are computed, and these values guide the selection of suitable components for signal reconstruction. Finally, the WTD technique is applied to the reconstructed signal to further enhance the denoising level. We applied this algorithm in both simulations and experimental analyses, comparing its performance against three commonly used methods: wavelet filter, Savitzky-Golay (S-G) filter, and CEEMDAN-WTD filter. The results indicate that the proposed algorithm significantly suppresses noise in the QEPAS gas detection system, enhancing the signal-to-noise ratio (SNR) from 105.25 to 595.85. Additionally, it achieves a correlation coefficient (R2) of 0.999 and a minimum detection limit (MDL) of 1.01 ppm. The findings highlight how the proposed approach could improve QEPAS gas detection precision, while also offering new insights for research and applications in related fields.
Read full abstract