Research On Optimal Design Of Spectroscopic Absorption Multi-species Trace Gas Detection System

With the development of economy and social progress,air pollution is becoming more and more serious.It is a long-standing problem to detect complex gases with high precision in harsh environment.In gas detection technology,spectral analysis has developed rapidly because of its advantages of high sensitivity,good repeatability,simple operation process and long service life.In the spectral gas detection system,for the spectral absorption detector,the effective absorption optical path of the gas chamber is one of the key parameters to determine the detection sensitivity of the system.Hollow core photonic crystal band gap fiber(HC-PBGF)was used as the gas chamber to obtain a gas detection system with high accuracy and high stability by absorption spectroscopy.HC-PBGF can realize the full effect of gas and light,and has the advantage of low bending loss.As a gas chamber,HC-PBGF has the advantages of simple structure,strong anti-interference ability,stable structure,and sufficient interaction between light and gas.However,in the application of system,HC-PBGF is difficult to coupling with ordinary optical fiber and the gas diffusion is slow.In addition,the current absorption gas detection system,mostly used for infrared gas detection,can not achieve the detection of ultraviolet absorption gas and fluorescent gas.On the one hand,due to the limitation of light source band,on the other hand,there are differences between absorption spectroscopy and fluorescence spectroscopy in the gas chamber structure,resulting in a single detection principle and limited detection types.In this paper,the application of HC-PBGF in gas detection system by absorption spectroscopy and the single detection principle were studied.In addition,the general average Empirical Mode Decomposition(EEMD)combined with wavelet algorithm was proposed to preprocess the noise reduction of gas experimental data,and the machine learning algorithm was used to determine the concentration of gas detection system.The main research work of this paper are as follows:(1)Based on the theory of gas molecular spectroscopy and energy level transition,the broadening mechanism of molecular absorption line was analyzed,and the environmental factors such as temperature that affect the width of gas spectral line are studied and determined.The characteristics of gas absorption spectrum and fluorescence spectrum were analyzed and studied.The relationship between spectral intensity and gas concentration was determined,which provides theoretical basis for the design of gas detection system.(2)An absorption gas detection system based on HC-PBGF was constructed.The system used the distributed feedback laser and the wavelength modulation technology to realize the modulation of the light source.The difference harmonic detection of methane was realized by using the HC-PBGF as the gas chamber and the optical attenuator as the reference light path.In the overall design of the gas chamber,the coupling device based on the casing method and the diffusion structure based on the differential pressure method were designed respectively,which solves the problems of large coupling loss between HC-PBGF and single-mode optical fiber in the system and too slow gas diffusion in the air core.The methane gas was used as the test gas,and the methane gas concentration and system stability were tested.The experimental results verified the feasibility of the system design.(3)In order to solve the problem of single gas detection principle and limited gas detection type due to the limitation of light source band and gas chamber structure in the gas detection system of absorption spectroscopy,the gas detection system based on multi-sensor principle was designed by combining the two gas detection methods of absorption spectrum and fluorescence spectrum is proposed.In the system,a light path system composed of light source,sensing optical fiber and gas chamber were designed,which realized the simultaneous detection of gas infrared absorption spectrum,ultraviolet absorption spectrum and fluorescence spectrum.Two typical gases,methane and sulfur dioxide,were used as test gases.The infrared absorption spectrum of methane,ultraviolet absorption spectrum of sulfur dioxide and fluorescence spectrum were tested.The test results of the system verified the feasibility of the design scheme.(4)Aiming at the problems of low gas concentration and weak signal energy in the gas detection system,and the detection signal is easy to be drowned by noise,an algorithm combining EEMD and wavelet method was put forword to reduce the noise of the gas measurement results and achieve the effective extraction of useful signals.For the determination of gas concentration,the traditional least square fitting method has the measurement error introduced by the incomplete linearity.In this paper,the improved algorithm of chicken flock optimization support vector regression machine was proposed to be used for the determination of gas concentration,the improved chicken colony optimization algorithm was used to optimize the parameters of the SVR,so as to improve the accuracy of gas concentration determination and the performance of gas detection system.