Gas Detection Technology And Application Based On Infrared Absorption Spectroscopy

The subject of this paper is derived from the National Key Research and Development Project:Research on the test method of operation reliability and safety in complex environment（project number:2016YFD0700101）.Compared with electrochemical,catalytic combustion,semiconductor and other gas sensors,gas sensors based on infrared absorption spectroscopy have the advantages of good selectivity,high measurement accuracy and short response time,and are widely used in environmental monitoring,combustion diagnosis and biomedicine.Non-dispersive infrared（NDIR）spectroscopy and photoacoustic spectroscopy（PAS）gas sensors have the advantages of small size,low cost and portability,making them easier to promote.A carbon monoxide（CO）,carbon dioxide（CO₂）gas sensor based on the NDIR principle and a miniaturized CO₂ sensor module were developed to meet the requirements of low-cost,low-power,anti-environmental interference and portable gas detection in complex environments.Photoacoustic spectroscopy gas sensing technology based on non-resonant photoacoustic cell and resonant photoacoustic cell was studied.Aiming at the safety requirements of agricultural operations under complex and open conditions,an early fire detection and location method was proposed,and fire detection application was carried out.Firstly,a mid-infrared CO and CO₂ detection system based on NDIR principle was developed.A broadband thermal light source with a radiation wavelength of 2μm–20μm and a modulation frequency of 4 Hz was used.In order to ensure the good selectivity of the sensor,two dual-channel pyroelectric detectors with central transmission wavelengths of 4.26μm/3.95μm and 4.66μm/3.95μm were used to detect CO₂ and CO,respectively.In order to match the broadband thermal light source with divergence angle of 15°and improve the sensitivity of the sensor,a closed multi-pass gas cell with a cavity length of 30 cm,a radius of 4 cm and an effective optical path of about 180 cm was designed.In order to reduce the impact of dust and moisture on the optical path,a gas pretreatment module composed of a filter and a drying tube was developed.The developed sensor system was integrated and its performance was tested.The detection ranges of the sensors are:CO₂,200 ppmv–1600 ppmv（parts per million in volume）and CO,0 ppmv–200 ppmv.When the gas flow rate is 1000 ccm,the response time of the system is about 30 s.The long-term stability of the sensor was tested,and the Allan variance was used to evaluate the limit of detection（Lo D）of the system.The results show that when the average time is 0.25 s,the Lo Ds of CO₂and CO are 5.66 ppmv and 0.94 ppmv,respectively.On this basis,in order to further reduce the volume and power consumption of the sensor,a miniaturized gas sensor module with a size of 80×50×50 mm was developed.An optical path with an optical path length of 60 mm was designed,and the optical path performance was simulated using Trace Pro software.The hardware circuit and software program based on STM32 microcontroller were designed.The experimental results show that when the average time is 1 s,the Lo D of the sensor is 0.45 ppmv.Secondly,a photoacoustic spectroscopy gas detection system based on non-resonant photoacoustic cell was developed.The generation process of photoacoustic signal was described from the aspects of light absorption and sound excitation,and the system structure was design.The filters with the central wavelength of 4.26μm and 3.4μm were selected to convert the radiation of the broad-band light source into narrow-band light.The size of the ultra-low noise microphone used is only 4×3×1.2mm.A chopper controller based on STM32 microcontroller was developed.The performance of the controller is basically the same as that of commercial instrument,and its size is smaller and the cost is lower.A non-resonant photoacoustic cell with a diameter of 25 mm and a length of 20 mm was developed.Through experiments,the optimal modulation frequency of the light source was 40 Hz.The performance of CO₂and methane（CH₄）was tested by the developed PAS sensor system.When the average time was 1 s,the Lo Ds of CO₂ and CH₄ were 1.41 ppmv and 1.61 ppmv,respectively.In order to increase the optical power of the cavity,a flat concave spherical mirror was used instead of a plane window in the optical outlet hole of the photoacoustic cell,and the signal-to-noise ratio（SNR）of the system was increased by 1.12 times.Then,two photoacoustic spectroscopy gas detection systems based on resonant photoacoustic cell were developed.（1）A gas detection system based on tunable laser and resonant photoacoustic cell was developed.The tunable laser used has a wavelength tuning range of 1520 nm to 1580 nm,and can detect multi-component gases such as CO₂（1572.34 nm）,ethyne（C₂H₂,1530.37 nm）,and ammonia（NH₃,1532.65 nm）through tuning the wavelength.In order to improve the sensitivity of the sensor and realize the dynamic gas detection,a resonant photoacoustic cell was designed.The photoacoustic cell was modeled and simulated to obtain a resonant frequency of 1612 Hz.The optimal modulation frequency of the light source was determined to be 806 Hz through experiments.The performances of CO₂,C₂H₂ and NH₃ were tested using the developed sensor system,and the fitting relationship between the concentration and the photoacoustic signal amplitude was obtained.It was proved by experiment that the photoacoustic signal amplitude was positively correlated with the optical power within a certain range of optical power.When the average time was 3 s,the Lo Ds of CO₂,C₂H₂ and NH₃ were 0.75 ppmv,0.01 ppmv and 0.43 ppmv respectively.（2）A gas detection system based on distributed feedback laser and resonant multi-pass photoacoustic cell（MPPC）was developed.A time division multiplexed CO₂/CH₄ gas sensing system was designed,and 1572.34 nm and 1653.73 nm were selected as the gas absorption lines.To increase the sensitivity of the photoacoustic sensor system,a resonant MPPC was designed.The parameters of the photoacoustic cell were optimized.The length and radius of the optimized resonator were 140 mm and 5 mm respectively,and the length and radius of the buffer cavity were 70 mm and 20 mm respectively.According to the geometric dimensions of the photoacoustic cell,a multi-pass gas cell was designed.The two concave mirrors of the multi-pass gas cell have a focal length of 100 mm,a diameter of 0.5 inch and a spacing of 150 mm,and the measured mirror reflectance is >95%.The distance between the two concave mirrors was determined to be 150 mm.By optimizing the central position of the incident hole and the incident angle,24 times of reflection were obtained.The photoacoustic cell was modeled and simulated,and the first-order resonant frequency obtained from the simulation was 1184 Hz.The optimal modulation frequency of the light source was determined to be 583 Hz through experiment.CO₂ and CH₄ were detected by the developed sensor system,and the fitting relationships between concentration,optical power and photoacoustic signal amplitude were obtained respectively.The SNR of the system was evaluated,the 1σ（σ is the standard deviation）values of CO₂ and CH₄ were 36.23 ppmv and 13.89 ppmv,respectively,and the corresponding normalized noise equivalent absorption coefficient（NNEA）were 7.13 × 10^-9 cm^-1·W·Hz^-1/2 and 2.29 × 10-9 cm^-1·W·Hz^-1/2,respectively.Through the analysis of CO₂ and CH₄ in coal mines,it is proved that the developed sensor system was feasible for coal gas analysis.The performance of single-pass photoacoustic cell（SPPC）and MPPC was compared.The experimental results show that the photoacoustic signal amplitude of the MPPC was about 10 times higher than that of the SPPC,which is slightly lower than the theoretical power enhancement factor,mainly considering the transmission loss of light in multiple reflections.By comparing the NNEA of the developed sensor system with other sensor systems,it was proved that the developed sensor system had good performance.Finally,the methods of fire detection and location for agricultural operation safety were studied based on the developed NDIR sensor system.In order to improve the stability of the system,a damping device were developed,which reduced the intensity of external vibration.Aiming at the impact of large temperature difference between day and night on the performance of the sensor system,a temperature control module was designed to ensure that the sensor can operate normally within the range of –40 ℃ to 85 ℃ by controlling the internal temperature of the sensor.According to the change amplitude and change rate of CO concentration,an adaptive early fire identification algorithm was proposed,which can avoid false alarm and missed alarm.The algorithm can automatically adapt to the change of CO gas concentration in the background environment.Through experiments,the threshold values of the change amplitude and change rate of CO concentration were determined to be 100 ppmv and 2 ppmv/s,respectively.Although CO₂ is an important reference index in the early stage of fire,it is extremely susceptible to interference from external factors and is not suitable to be used as the standard of fire judgment alone.A multi-parameter fusion fire detection method was proposed based on the change correlation between CO₂ and CO,which normalized the concentration ratio of CO and CO₂,that is,using C（CO）/C（CO₂）as the alarm signal,and the alarm threshold was determined to be 0.1 through experiments.Based on the gas turbulent diffusion model,combined with particle swarm optimization algorithm,a mobile early fire location method was proposed,and a field fire source location experiment was carried out on the campus of Jilin University.The experimental results proved the feasibility of the location algorithm for fire location.The developed sensor system was installed on the harvester to carry out the simulation fire detection experiment.The experimental results showed that the developed sensor system was suitable for agricultural fire detection and early warning.The main innovations of this thesis are as follows:1.A resonant multi-pass photoacoustic cell was designed to improve the sensitivity of the sensor system.Compared to a single-pass photoacoustic cell,the photoacoustic signal amplitude in the multi-pass photoacoustic cell was increased by about 10 times.2.According to the variation amplitude and rate of CO concentration,an adaptive early fire identification algorithm was proposed.Simultaneously,according to the change correlation between CO and CO₂ concentration in early fire,a multiparameter fusion early fire detection method based on normalized CO to CO₂ concentration ratio was proposed,which improves the accuracy of early fire detection.3.Based on the gas turbulent diffusion model and particle swarm optimization algorithm,a mobile early fire location method was proposed.The feasibility of this method was proved by using the developed NDIR sensor.