Error Processing of Real-time Coal Mine Oxygen Detection Data Based on Wavelet Threshold Function Denoising

ZHANG Huaqian; YU Qing

doi:10.16545/j.cnki.cmet.2023.06.007

Abstract：The oxygen detection system based on Tunable Diode Laser Absorption Spectroscopy(TDLAS)technology can accurately detect the oxygen volume fraction in the underground environment of coal mines.However,when using this system to detect oxygen volume fraction,it is often affected by noise signals such as shot noise,thermal noise,and Gaussian white noise,resulting in significant errors in oxygen volume fraction detection.In order to reduce the impact of noise on oxygen volume fraction detection data,firstly,polynomial smoothing algorithm was applied to perform high and low frequency analysis on the oxygen volume fraction detection curve.Secondly,a method for processing errors in coal mine oxygen detection data based on wavelet threshold function denoising was proposed.The denoising effects of Daubechies wavelet and Symlets wavelet were compared,and the best denoising wavelet basis function was selected based on signal-to-noise ratio and root mean square error.Finally,the stability of oxygen volume fraction detection before and after noise reduction was compared by calculating the standard deviation and maximum amplitude of the oxygen volume fraction detection curve.The experimental results showed that the sym6 wavelet basis function had the greatest correlation with the second harmonic signal.Applying this wavelet to denoise the second harmonic signal improved the signal-to-noise ratio by 16.785dB.Before and after denoising,the maximum amplitude of the 20％oxygen volume fraction detection curve decreased by 0.04％,and the standard deviation decreased by 0.029 5,significantly improving the stability of oxygen volume fraction detection.