Research On Inversion Method Of Atmospheric Bioaerosol Detection Data For Fluorescent Lidar

Bioaerosol refers to biomass(microorganisms,plants or animals)suspended in the air and is a sub-category of particles released from the terrestrial and marine ecosystems into the atmosphere.Bioaerosols in the atmosphere are closely related to human health,and bioaerosols have been f'ound to cause certain human diseases such as tuberculosis,Legionnaires'disease an d different forms of bacterial pneumonia,coccidioidomycosis,influenza,measles and gastrointestinal diseases.This thesis introduces a laser-induced bioaerosol fluorescence lidar system for nighttime detection,in which 266nm is used as the excitation wavelength,the single pulse energy is 100mJ,the repetition frequency is 10Hz,and the SNR is based on the relevant system parameters.The perspective verifies the reliability of the system.The results show that when the SNR is 10 and the ozone concentration is Oppb,40ppb,80ppb,120ppb respectively,the detectable height of the system can reach 5km,3km,2km,1km respectively,so the system is considered to be capable of detecting atmospheric organisms.Based on the system,the data inversion scheme is further discussed.This thesis proposes to use the FastICA algorithm to separate the original received echo spectra and verify the original echo spectra obtained by the simulation.The verification group took two groups,which were biological substances containing biological components and non-bioaerosols(Casel)and substances with fluorescence effects containing bioaerosols of different growth stages of the same kind of bacteria(Case 2).Comparing the spectral patterns and true values obtained by FastICA,it is found that ICA has the ability of resolution for both sets of spectra.Based on the genetic algorithm and the fluorophore library information,the standard mixed fluorescence spectra were identified and matched.The relative error of the matching coefficients of the two sets of standard spectra are:6.7%,0.0%,30.0%,12.5%and 0.0%,0.0%,0.0%,1.0%.The relative error results indicate that the genetic algorithm has the ability to distinguish the specific types of fluorescent substances and determine the proportion of specific components.Then,the cascade of FastICA and genetic algorithm is simulated.Combined with the genetic algorith m an d the fluorescence spectrum library,th e mixed fluorescence spectra af ter amplitude correction are matched.The coefficient matching errors of the two sets of spectra were;4.1%,0.0%,5.1%,13.2%and 1.1%,5 1.9%,21.3%,97.8%.The results show that the p roposed algorithm can achieve spectral decomposition,amplitude correction,species iden tification and coefficient matching for Casel and Case2.It has certain theoretical value for the later experimental detection.