Research On SO₂ And NH₃ Mixed Gas Detection Algorithm Based On DOAS Technology

Air pollution brings great harm to the ecological environment.The harmful gases produced by coal-fired power plants,such as SO₂,NH₃,particulate matter,etc.,are one of the main sources of air pollution.As a result,our country has formulated a series of environmental protection policies,which has also accelerated the development of online monitoring instrument for coal-fired power plants.This thesis developed an online monitoring system for SO₂ and NH₃ based on differential optical absorption spectroscopy technology.For SO₂ and NH₃ in the range of 7～80 ppm,this thesis chooses a more sensitive absorption band to solve the problem of difficult detection of low-concentration SO₂ and NH₃.However,SO₂ and NH₃ have spectral aliasing in the ultra-sensitive band,and the concentration inversion will interfere with each other.Therefore,the thesis mainly studies the noise processing in the absorption spectrum of SO₂ and NH₃,the mixed gas spectrum separation technology and the mixed gas concentration inversion algorithm.（1）Optimizing the absorption bands of SO₂ and NH₃.Based on the principle of strong absorption of the target gas and weak absorption of the interfering gas,the optimal inversion bands were selected for the two gases.The absorption band of SO₂was 280～310 nm,and the absorption band of NH₃ was 201～214 nm.The data was directly obtained from the spectrometer contains additive noise and multiplicative noise.This paper analyzes the characteristics of these two kinds of noise,selects a suitable filtering algorithm,and achieves the purpose of filtering noise.（2）Calculating the differential absorption cross section of SO₂ and NH₃.Gas absorption includes two parts:fast change and slow change.Only in the fast change can there be gas concentration information.Therefore,it is necessary to separate the slow change from the absorption spectrum.In this article,the traditional polynomial filtering method is used to obtain the slow change in the absorption spectrum,the fast change and the slow change are separated by a differential technique,and the differential absorption cross section of SO₂ and NH₃ is obtained by using a calibration gas.（3）Studying the concentration inversion algorithm of single gas.The least squares and Fourier transform methods are used to carry out the concentration inversion,and the nonlinear absorption problem arising from the least squares method is corrected.The results show that the absolute values of the relative errors of SO₂ and NH₃concentrations calculated based on the Fourier transform method are both less than 3%;when using the least square method,the absolute value of the relative errors of SO₂inversion is less than 1%,and the absolute value of NH₃ is less than 5%.（4）Studying the separation technology of the characteristic spectrum of the mixed gas and the concentration inversion algorithm.In view of the cross-sensitivity of SO₂and NH₃,this thesis proposes the characteristic spectral separation of SO₂ and NH₃based on the segmentation method and the frequency domain method of Fourier transform.The results show that:the absolute value of the relative error of SO₂ reversed by the subsection method is less than 2%,while the absolute value of the relative error of NH₃ is less than 6%;the absolute value of the relative error of SO₂ reversed by the Fourier frequency domain analysis method is less than 6%,the overall error of NH₃ is less than 5%.