A Novel Calibration-free Differential Optical Absorption Spectroscopy And Its Application In NO Measurements

Nitrogen oxide（NO_x）is a major chemical/combustion emission pollutant and is a binding emission monitoring indicator required by the state.NO is prone to form other NO_x through photochemical reactions,which have a significant impact on acid rain,ozone depletion and the greenhouse effect.As more than 95%of NO_x exists in the form of NO,real-time online monitoring of NO pollutants is of great significance as environmental requirements become increasingly stringent.Differential optical absorption spectroscopy（DOAS）technology has the advantages of simple structure,high accuracy,and multi-component real-time detection,and is widely used for the detection of flue gas component concentrations in highly polluting and energy-consuming industries such as chemical plants and coal-fired power plants,and is also the mainstream method for online monitoring of NO gas concentrations at present.This paper addresses the key issues of calibration-free UV Differential optical absorption spectroscopy in the field of gas concentration detection.The selection of the optimal measurement band for NO measurement system and the principle of temperature nonlinearity compensation for a wider range of applications are investigated to provide theoretical guidance for the wide application of UV Differential optical absorption spectroscopy.The variation of UV light source characteristics is studied,and a DOAS inversion algorithm based on light intensity calibration-free is proposed to achieve accurate measurement of gas concentration by real-time correction of UV light source light intensity variation.The effectiveness of the method is verified by comparing and analyzing the concentration measurement results and the Allan variance test of zero gas.The main conclusions of this paper are as follows.（1）An improved guided filtering algorithm is proposed to realize the pre-processing of spectral data,which ensures high measurement accuracy while suppressing noise.（2）To address the problem that the variation of light source intensity in DOAS system affects the stability of gas measurement,a new light intensity calibration-free DOAS inversion algorithm is proposed in this paper,which uses the domain broadband cross section to normalize the narrowband cross section of gas characteristic absorption,so as to obtain the equivalent absorption intensity parameterα_eq（λ_k）that is not affected by the variation of light source intensity,and by comparing the The gas concentration can be deduced by comparing the measured value of this parameter with the calculated value of the standard cross section.Compared to conventional algorithms,this algorithm does not require complex calculations such as polynomial fitting and signal filtering,and is easier to implement in hardware.The results show that the characteristic absorption peaks of NO only appear at four wavelengths,such as 195.5 nm（absorption peak 1）,204.7 nm（absorption peak 2）,214.8 nm（absorption peak 3）and 226.2 nm（absorption peak 4）,and the half-height and full width of these four absorption peaks are all about 1 nm.1 nm.The linear regression coefficient R2 of the characteristic absorption peak 4（226.2 nm）reached a maximum of 0.99817 using the algorithm of this paper,which verified the feasibility of the new algorithm.（3）The effects of nonlinear absorption and temperature on the NO absorption spectrum were investigated,and the NO nonlinear compensation function and the temperature interference compensation function were established.The experimental validation was carried out at different NO proportional concentrations and different temperatures,respectively.（4）The field tests in the actual production environment of the chemical plant showed that the measurement deviation of the self-developed DOAS measurement system was no more than 5.25%compared with the commercial instrument;the Allan ANOVA of the zero-gas measurement data showed that the proposed new method has a lower detection lower limit compared with the traditional DOAS algorithm based on baseline fitting.