Tunable Diode Laser Absorption Spectroscopy In Atmospheric Monitoring And Combustion Diagnostics

In recent years,Tunable Diode Laser Absorption Spectroscopy(TDLAS)has attracted much attention in quantitative measurements due to its good selectivity,high sensitivity,real-time,and online measurements,and has shown great promise for applications in atmospheric monitoring and combustion diagnostics.With the increasing demand for atmospheric environment detection and combustion flow field diagnosis,it is important to develop high-sensitivity,high-precision and fast TDLAS measurement methods and to develop TDLAS-based multi-parameter measurement techniques for combustion fields.TDLAS is mainly divided into Wavelength Modulation Spectroscopy(WMS)and Direct Absorption Spectroscopy(DAS).In this paper,we investigate the application of TDLAS to atmospheric monitoring first,focusing on the optimization of the sensitivity and accuracy of WMS concentration detection and the concentration inversion method for open-path calibration-free WMS.In order to improve the WMS concentration detection sensitivity,a double-enhanced multi-pass cell was designed based on the conventional Herriot multi-pass cell,which can double the effective optical pathlength with the same size and improve the mirror utilization and detection sensitivity.An absorption pathlength of 73.926 m was achieved at a base length of 33.3 cm.Measurement accuracy was also optimized by the temperature control of absorption cell,wavelength locking and light intensity normalization.The system was used for long-term monitoring of atmospheric CH4 concentrations and achieved a detection precision of 10 ppbv.To compensate for the in situ measurement limitations of the closed-path system,research on open-path WMS was also carried out.A calibrationfree WMS concentration fitting method based on software lock-in amplifiers was developed and used for atmospheric NH3 concentration detection in wheat fields to address the difficulties in calibration of open-path WMS.In addition,research on TDLAS-based multi-parameter measurements in combustion diagnosis was carried out.DAS-based temperature,pressure,species concentration and combustion field distribution measurements were carried out using two absorption lines of 7444.35 cm-1 and 7185.6 cm-1 for H2O and 4993.75 cm-1 for CO2.In order to reduce the influence of the DAS baseline fitting error,an inversion method based on the derivative of absorbance was investigated and the feasibility of this method was verified by simulations and experiments.In order to improve the precision and speed of inversion for temperature,pressure,and concentration,and to reduce the dependence of DAS measurements on spectral line parameters,a transfer learning-based inversion method was investigated for the H2O absorption lines as an example.A two-color pressure measurement method was also proposed to address the problem that pressure inversion is influenced by concentration,and the average inversion errors of experimental measurement results were within 5%.As single-path TDLAS measurements can only provide line-averaged measurement results,DASbased combustion field distribution measurements were also carried out.Measurements and reconstructions of the temperature and concentration fields of the flat-flame burner and the exhaust plume of a miniature turbojet engine were carried out using limited projections of 10 and 12 rays.In order to reduce the DAS baseline fitting errors due to signal fluctuations caused by turbulence in the combustion flow field,a method for recovering the absorbance from the first order derivative of the absorbance was investigated in combination with a deep learning approach,with the potential to improve the quantitative accuracy of the temperature and concentration field reconstructions.This paper focuses on the study of the application of TDLAS in atmospheric monitoring and combustion diagnostics.Based on WMS,measurement systems for atmospheric CH4 and NH3 were established and optimized,providing experimental basis and technical support for atmospheric trace gas monitoring.Combined with deep learning and derivative spectroscopy,DAS-based inversion method for combustion flow field temperature,pressure and concentration,and combustion field distribution measurement method were investigated,which improves the accuracy,inversion speed and noise immunity of multi-parameter measurements and provides theoretical and experimental basis for combustion diagnosis.