Sulfur Dioxide Detection Signal Denoising Based on Support Vector Machine
Zhifang Wang,
Shutao Wang
Issue:
Volume 3, Issue 4, December 2018
Pages:
54-60
Received:
22 January 2019
Accepted:
26 February 2019
Published:
19 March 2019
DOI:
10.11648/j.jeece.20180304.11
Downloads:
Views:
Abstract: A system for detecting sulfur dioxide (SO2) based on differential optical absorption spectrometry theory was studied. The detection system can eliminate the noise from light source and light path by using the double optical path. Background noise was generated by the photoelectric device. It also effects the quantitative analysis. The Support Vector Machine (SVM) is proposed to process the SO2 ultraviolet absorption spectrum. The SO2 ultraviolet absorption spectra at 220nm-340nm were obtained by using the SO2 detection system in this article. Then the spectral was denoised by the SVM. The experimental results showed that the absorption line was more smoothness after denoising by the SVM, and the SNR and mean square error were 48.9398 and 1×10-7, respectively. The de-noising data was applied to the SO2 detection system, the linearity of the measurement was good with the coefficients of more than 0.9971. Compare the result with the wavelet and Empirical Mode Decomposition (EMD) denoising methods, which illustrates that SVM has better effects. It shows that the SVM method applied to noise reduction of SO2 detection system is superior.
Abstract: A system for detecting sulfur dioxide (SO2) based on differential optical absorption spectrometry theory was studied. The detection system can eliminate the noise from light source and light path by using the double optical path. Background noise was generated by the photoelectric device. It also effects the quantitative analysis. The Support Vecto...
Show More
