A Gas Analysis Model with Uninformative Variable Elimination Using Nonlinearity Compensation by RBFNN

Author(s):  
Panpan Zhang ◽  
Yanbin Zhang ◽  
Hui Cao ◽  
Zonglin Ye ◽  
Dapeng Yan ◽  
...  
2011 ◽  
Vol 63-64 ◽  
pp. 878-881
Author(s):  
Yong Li ◽  
Yong Sun ◽  
Lian Cong Wang

Coal mine underground environment gases gas composition is very complicated, especially coal self-ignition different oxidation stage symbol of mine gas and release of period of reckoning produce of the coal mine enterprises poisonous and harmful gas production safety, caused great harm to the mine production period and needed after the reckoning mine air composition for monitoring. Existing gas chromatographic analysis is slow and not on-line analysis problem, sensor use short-lived, need often set-up insufficient, we put forward such a kind of brand-new based on the Fourier transform infrared spectrometer (FTIR) in-well environment gas analysis method. Combining with gas absorption spectral characteristics and analysis from spectrometer parameter selection, request the sample gas preparation, feature extraction and analysis model is established in variable introduces a set of effective coal mine underground environment gas spectral analysis method, and then gives C2H6, C3H8, CH4, i-C4H10, n-C4H10, C2H4 C3H6, C2H2, SF6, CO and CO2, the test results. The results showed that the FTIR spectrometer in spectral resolution for 1cm-1, optical path for 10cm circumstances, all gas resolution than 2x10-6, meet mine environment gas detection requirements. This shows that the infrared spectrum analysis is a potential mine environment gas analysis, this paper party could realize technology of coal mine underground environment gas on-line analysis.


Sign in / Sign up

Export Citation Format

Share Document