Research On TDLAS Spectral Signal Processing Technology Based On Particle Swarm Optimization And Theory Of Spectral Line Shape

In order to improve the detection accuracy of Tunable diode laser absorption spectroscopy（TDLAS）,an improved particle swarm denoising algorithm that improves the signal-to-noise ratio of the absorption spectrum.The influence of the measured gas temperature to the spectral shape is studied,and a temperature adaptive concentration compensation algorithm based on the theory of spectral line shape is designed.（1）Using improved particle swarm optimization to improve the signal-to-noise ratio of TDLAS signals.Aiming at the problems of the back-end signal processing circuit is simple and the spectroscopy signal is disturbed by noise seriously in TDLAS direct absorption spectroscopy.An improved particle swarm denoising algorithm is proposed to improve the signal-to-noise ratio of original spectral signals.To verify the feasibility of the algorithm,CH₄ gas was selected as the experimental gas,and build the CH₄ concentration measurement system based on TDLAS,along with a back-end data acquisition circuit.The main structure of the improved particle swarm denoising algorithm has two parts,the first part is to determine the fitness function based on the line-shape function of the original spectral signals,which is determined by the theory of spectral line shape.Using the line-shape function as the fitness function can accelerate the convergence speed of the algorithm.The second is to set three learning factors and fitness values to specifically remove system white noise and interference fringes from the gas absorption spectrum.Verifying the performance index of this algorithm,the processing results of different algorithms are compared,and the signal-to-noise ratios output by several processing methods are calculated respectively.For verifying the detection accuracy and stability of the improved particle swarm denoising algorithm,repetitive experiment and calibration experiment are conducted.This algorithm exhibits strong denoising performance,improves the signal-to-noise ratio of the original spectral signals,the signal-to-noise ratio（SNR）is improved by 4.17 times,and the minimum detection limit in the experiment is15.3 ppb.R²=0.9999 is calculated in the calibration experiment,and the error is less than0.1 ppm in the repeatability experiment of constant methane at 2 ppm concentration.（2）A temperature adaptive concentration compensation algorithm based on the theory of spectral line shape.Aiming at the problem of large errors in measuring the gases at different ambient temperatures.A temperature adaptive concentration compensation algorithm based on the theory of spectral line shape is proposed.In the actual measurement of trace gas concentration,the gas temperature to be measured varies continuously with the ambient temperature,and the line intensity changes with temperature.In addition,the half width at half maximum of the gas absorption spectrum also presents a correlation relationship with temperature.Based on the HITRAN database,the relationship between gas absorbance and half peak half width at different temperatures is simulated and calculated to verify the feasibility of this method.In actual experimental testing,measuring absorbance and half peak half width of standard gas at different temperatures in advance and then establishing a temperature compensation model can realize adaptive compensation of gas concentration at different gas temperatures.To verify the performance of this compensation method,place the standard gas at different temperatures and perform compensation calculations in this way.In the verification experiment,the relative concentration error of methane compensated for different temperatures is less than 11%.Experiments showed that the accuracy of the calculated gas concentration after compensation was significantly improved.The proposed method provides a feasible scheme for measuring extreme gas concentration.