Research On Identification Method Of Atmospheric Bio-aerosol Components Based On Fluorescence Spectra

Bioaerosol is mainly composed of bacteria,viruses,dust mites in the atmosphere,etc.It is a special material with more life activity,mainly distributed in the atmospheric troposphere.The spread and spread of bio-aerosol in the air may cause the spread of epidemic diseases between humans and between humans and animals,which is more harmful than non-biological aerosols.In order to effectively prevent the biological aerosol in the atmosphere from harming the human body,it is necessary to identify the type of biological aerosol and detect its concentration.Lidar has high spatial and temporal resolution and high measurement accuracy,coupled with the interaction of laser and bioaerosol particles,making it one of the important means of bioaerosol observation in the atmosphere.In this paper,based on fluorescence spectroscopy,the composition identification of atmospheric biological aerosols is studied.Excitation of organic biological components contained in bio-aerosols by ultraviolet-wavelength laser induction(266 nm wavelength),to obtain effective information of fluorescence spectrum,and to realize the identification of the main components of bio-aerosols in the atmosphere.Because different types of bioaerosol particles have different characteristics of fluorescence spectra,the range of the characteristic peaks of the excited bioaerosol particles is different,which can be used as the basis for component identification.Based on atmospheric echo spectroscopy,a method is proposed.The independent component analysis algorithm with blind source separation function can realize the preprocessing processes such as noise removal,baseline drift,Raman echo spectroscopy,etc.,thereby obtaining mixed fluorescence spectra.It is known from the results that the extracted Raman spectrum and the mixed fluorescence spectrum are compared with the original spectrum,which proves that the independent component analysis algorithm can extract the mixed fluorescence spectrum.Finally,the genetic algorithm is used to extract the single fluorescence and obtain the relative mass concentration of the mixed fluorescence spectrum to realize the identification an d detection of biological components.In this paper,the fluorescence component spectra of different types of bioaerosol particles are identified for different ratios.Spectral decoupling of 3,4,and 5 different kinds of components is performed respectively.The relative errors of the decoupled spectral types of the obtained components are 12%,5%,20%.The results show that when the algorithm decouples four biological components,the relative error is small,the accuracy is high,and the component analysis is reasonable.In order to further verify the reliability of the algorithm,this paper also optimizes the population size of the decoupling algorithm.The results show that the change of the population size has little effect on the algorithm results;by changing the convergence criteria of the algorithm,we can see that the convergence condition is that the fitness difference is less than In the case of 0.05,the optimal solution can be obtained,and the relative error is controlled within 3%,which can meet the decoupling requirements.The successful application of the algorithm can improve the detection capability of fluorescent lidar,and is expected to achieve real-time online identification of atmospheric biological aerosols.