Application of a VMD-Wavelet Packet Analysis Denoising Method to Gas Detection Signals
To improve the detection accuracy of gas concentration in tunable semiconductor laser absorption spectroscopy,an optimized variational mode decomposition combined wavelet packet denoising algorithm,based on the genetic algorithm(GA),was proposed.The proposed algorithm can solve the electrical and optical noises of the second harmonic signal in gas detection.Initially,GA optimized the decomposition mode number K and penalty factor α,and the optimal parameter combination was obtained.Then,the second harmonic signal was decomposed by the optimal parameter combination,and a series of intrinsic modes(IMF)was obtained.Pearson correlation coefficient was used to select the pure and noisy signals,and the noisy signal was denoised using a wavelet packet.Finally,the denoised signal was reconstructed with the pure one to obtain the second harmonic signal after denoising.Results show that compared with other noise reduction algorithms,the proposed algorithm can effectively remove the noise in the second harmonic signal,reduce the influence of external noise on the detection process,and improve the detection accuracy of gas concentration.
NETL
NSTL
万方数据
