Title :
Extending AOTF-NIR Spectrometer to Gas Measurement
Author :
Huimin, Hao ; Mi, Pan ; Junhua, Liu
Author_Institution :
Xi´´an Jiaotong Univ., Xi´´an
Abstract :
As a new generation spectrometers, Acousto-Optic Tunable Filter (AOTF)-Near Infrared (NIR) spectrometer had been widely applied in many domains due to its remarkable advantages such as higher resolution, faster and more robust. But, to date, the availability of AOTF-NIR spectrometer is limited to liquid and solid detection. In this paper, the application of AOTF-NIR spectrometer in gas quantitative determination is explored. A vacuumable gas cell was designed and assembled to Luminar 5030-731 AOTF-NIR spectrometer for obtaining spectra of gas samples including methane (CH4) and carbon monoxide (CO), and its mixture. All detected gas samples of different concentrations are prepared dynamically by diluting standard gas with super pure nitrogen via precision volumetric gas flow controllers. To test the performance of refitted spectrometer, gas samples were injected into the gas cell and the spectra of wavelength between 1100 nm-2300 nm were collected. The relative absorbency of CH4 of different concentration was compared. It shows obvious proportional increase along with the concentration increase. This indicates that the AOTF-NIR spectrometer mounted gas cell can be used for scanning gas samples and obtain exact spectra. And then, prediction model of unknown concentration gas mixtures of CH4 and CO was created by using kernel partial least squares (KPLS). The effectiveness of the model is tested by the samples of predict set. The mean square error (MSE) of the predicted gas is below 1.64% when the concentrations of component gas are over 0.40%. The exploratory work indicates a new way for gas detection and provides the base for developing of NIR gas detector.
Keywords :
chemical variables measurement; gas sensors; infrared spectrometers; least squares approximations; mean square error methods; acousto-optic tunable filter; gas detection; gas measurement; kernel partial least squares; luminar 5030-731 AOTF-NIR spectrometer; mean square error; near infrared spectrometer; precision volumetric gas flow controller; vacuumable gas cell; Acoustic signal detection; Assembly; Filters; Infrared spectra; Nitrogen; Predictive models; Robustness; Solids; Spectroscopy; Testing; Acousto-Optic Tunable Filter; Kernel Partial Least Squares; Near Infrared; Quantitative Analysis;
Conference_Titel :
Electronic Measurement and Instruments, 2007. ICEMI '07. 8th International Conference on
Conference_Location :
Xi´an
Print_ISBN :
978-1-4244-1136-8
Electronic_ISBN :
978-1-4244-1136-8
DOI :
10.1109/ICEMI.2007.4350473