Study On The Analytical Techniques For Tunable Diode Laser Absorption Spectroscopy

Tuned Diode Laser Absorption Spectroscopy(TDLAS)is widely used in gas detection due to its high sensitivity,high precision,good selectivity,strong anti-interference ability and ability to work at room temperature.The existing line analysis methods have some key problems in system parameter optimization,air pressure detection and components recognition of aliasing spectral line,which limits the application range of the technology.In view of the above problems,this thesis aims to improve the TDLAS component identification ability and detection performance,to carry out research on the tuned laser absorption spectrum line analysis technology.The main work is divided into the following three aspects:1.The influence of each modulation parameters on the frequency characteristics of the second harmonic signals is observed through experimental studies.Firstly,the influence of each parameter on signal line shape,frequency characteristics and introduced noise is obtained.Then the decision of multi-parameter joint change on the spectral line band is analyzed.Finally,the basic selection methods of each parameter based on frequency characteristics are summarized.The optimal second harmonic signal can be obtained by parameter selection in the premise of considering the system detection requirements and hardware condition limitations,which provided experimental basis and reference method of parameter optimization for the practical application of TDLAS technology.2.A mathematical model of spectral line broadening in harmonic detection is established.Firstly,the Voigt line type obtained by Gauss line type and Lorenz line type convolution is simulated as the ideal line.Gaussian fitting and Lorentz fitting are performed on the Voigt function,with the ratio of the fitting degree of Gauss and Lorentz as the eigenvalue.A mathematical model of the ratio of Lorentz to Gaussian fit and pressure is established by changing the proportion of FWHM of Gaussian and Lorentz functions.Secondly,the various factors affecting the absorption line type in the actual measurement are simulated.Such as laser linewidth,random noise of the laser,and background interference caused by etalon stripes in the optical system.Then,the model of the pressure and absorption line peak width,the ratio of the pressure to the absolute value of the second harmonic and the fourth harmonic peak and the ratio of the pressure to the second harmonic and the fourth harmonic peak height is established,and comprehensive comparison and analysis are carried out.Finally,the established spectral line broadening model was verified by the measured data.3.Based on the wavelet transform,an algorithm for spectral component identification is proposed.The time-frequency characteristic matrix of the signal is obtained by continuous wavelet transform of the second harmonic signal,and the correlation between the wavelength and the spatial frequency is analyzed by the time-frequency characteristic matrix of the gas to be measured and its mixture.From the wavelength dimension analysis,the characteristic absorption position interval of the gas to be tested in the mixed gas is extracted;then the spatial frequency dimension analysis is performed in the extracted interval to realize the gas component identification.The component identification is verified by the experimental data.The concentration calculation is carried out according to the characteristic absorption position of the extracted components in the mixture,and the component identification and the quantitative detection of the concentration are realized.