| In recent years,the demand for trace gas sensors in industry,agriculture,aerospace and other fields is increasing.Photoacoustic spectroscopy(PAS)gas sensing technology has the advantages of high sensitivity,fast response,good selectivity and large dynamic range.It is an ideal method to realize real-time monitoring of trace gases,and has important research value and wide application prospect.CO gas plays an important role in the fields of industrial applications,air pollution monitoring,indicating power system faults,and medical diagnosis.A trace CO gas detection system based on PAS is designed in this paper,it is combined with the second harmonic signal wavelength modulation spectroscopy technology(2f-WMS),differential photoacoustic cell(DPAC)and the denoising algorithm,and it realizes highly sensitive detection of CO gas.The main research content are as follows:The basic principle of 2f-WMS is discussed in detail,the broadening mechanism of gas absorption line and the corresponding line-shape are analyzed,and the harmonic expression are given for the absorption line shapes of Lorentzian signals,the amplitude and linewidths of the 2f signal have been calculated as a function of the modulation depth,which provides a theoretical basis for the selection of the modulation depth.The relationship between the photoacoustic signal and the resonance mode of the photoacoustic cell(PAC)is calculated and obtained,which provides a theoretical basis for the optimal design of the DPAC.On the basis of theoretical research,a high sensitivity CO-PAS sensor is developed.The sensor uses a near-infrared tunable fiber laser combined with an erbium-doped fiber amplifier(EDFA)as the system light source and uses a DPAC operating in the first-order longitudinal resonance mode to amplify the photoacoustic signal.In oeder to ensure that the maximum PAS signal amplitude is obtained without distorting the signal,the modulation current amplitude is optimized firstly and then the optimal parameter is used for subsequent experiments.The denoising algorithm is studied to further improve the detection performance of the CO-PAS sensor.In the PAS system,the gas absorption is a non-linear and nonstationary process with random and non-random noises.Therefore,The Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)algorithm is introduced to PAS field.The original signal is decomposed into several intrinsic mode function(IMF)components and residual component by CEEMDAN,then Savitzky-Golay(S-G)filtering algorithm were used to denoise each component,and then the signal-dominated component was adaptically selected to reconstruct signal.By analyzing the experimental results,the SNR of the system is increased by 4.6 times,the effectiveness and feasibility of the proposed algorithm in the PAS field are verified. |
