| The release of bioaerosol has become the main mode of bioterrorism,as well as the strong infection and rapid spread of high-risk pathogenic microorganisms carried by cross-border populations poses a threat to humans and society,the rapid monitoring of bacteria and other aerial biological species is becoming increasingly important.More and more countries have been working to effectively and quickly implement real-time monitoring and control methods for bioaerosols to provide faster and more accurate warnings and reduce unnecessary losses.In China,detection and identification technology for bioterrorism attacks is developed relatively late,and there are fewer testing instruments and means,which will affect the prevention of counter-terrorism security.Therefore,the research on bioaerosol detection methods has important practical and academic value for biosafety and bioterrorism prevention and terrorism.In this paper,Raman spectroscopy and fluorescence spectroscopy are used to identify and analyze the collected aerosol standard samples.Raman spectroscopy as the main means and fluorescence spectroscopy as the auxiliary means,taking 20common bioaerosol samples as the research object.Through the literature review and experiments,the Raman spectra of 262 samples of these 20 kinds of bioaerosol samples were collected,and the fluorescence spectra of some samples were collected and analyzed.The obtained data were preprocessed,and the classification of bioaerosol was realized by system cluster analysis,principal component analysis and support vector machine,which laid a foundation for the detection,classification and early warning of bioaerosol.This thesis mainly carried out research work on bioaerosol from the following aspects:1.The Raman spectra of bioaerosol samples were collected by confocal micro-Raman spectrometer.The best detection parameters were selected by analysis and comparison,including laser wavelength,integration time and confocal aperture.Taking pine pollen samples as experimental objects,the best acquisition parameters of bio-aerosol pollen were obtained by comparative analysis:laser wavelength785nm,integration time 30s,confocal aperture 300?m,scanning range 300-1800cm-1.Because bacteria and fungi can not directly obtain their Raman spectra by traditional Raman spectroscopy,so the surface-enhanced Raman scattering?SERS?technology is used for bacteria spectra collection,which uses"flower-like"nano-silver sol as the substrate and uses 532nm wavelength laser.Controlling the mixing volume ratio and binding time of the bacteria sample and the silver sol,the Raman signal intensity is enhanced when the laser power is attenuated by 100 times or 1000 times,and the Raman spectrum of bacteria and fungi is successfully collected.2.The Raman spectra of the collected bioaerosol samples were processed and analyzed.Firstly,the Raman spectrum was preprocessed.Through comparative analysis,the Savitzky-Golay method with a window size of 11 points was used for smoothing,the linear fitting method was used for the baseline,and the maximum normalization method was used for normalization.Eliminate noise,baseline drift,and magnitude differences in spectral acquisition as much as possible.For the processed spectral data,IBM SPSS Statistics 19 statistical analysis software was used to carry out systematic clustering analysis,and the complete Raman spectral data was used for classification,and the classification results were verified.The accuracy rate reached 91.67%,indicating that the system clustering method had certain effectiveness in the identification and classification of biological aerosol Raman spectra.Then the principal component analysis and support vector machine are used to establish the recognition model.With 182 sample data as the training set and 80 sample data as the prediction set,the optimal parameters established by the SVM model are selected by contrast analysis.The[-1,1]normalization and polynomial kernel functions have a recognition rate of 98.75%,which laid the foundation for the subsequent research on bioaerosols.3.Fluorescence spectroscopy was used to detect the fluorescence spectra of four bio-aerosol samples?Pan Pollen,Escherichia coli,Staphylococcus aureus,Candida albicans?,and the fluorescence peaks have carried on the comparative analysis,through the fluorescence spectra of four samples can be obvious:fluorescence spectrum can better classification identification of the sample.Fluorescence spectra of three strains with a concentration gradient of 10-1-10-88 mol/L were obtained by excitation of 279 nm wavelength excitation light and 289 nm wavelength excitation light.And the classification accuracy of the fluorescence spectra of the three kinds of bacteria(concentration 10-1mol/L)by principal component analysis can reach 100%.However,because the fluorescence spectrum acquisition instrument requires a larger amount of sample than the Raman spectroscopy acquisition instrument,due to the type and quantity limitation of the experimental sample,only four samples of the fluorescence spectrum are detected.In the subsequent work,it is necessary to further increase the number of samples detected by fluorescence spectroscopy to better illustrate the important role of fluorescence spectroscopy in bioaerosol research.In summary,Raman spectroscopy can effectively distinguish 20 kinds of bioaerosol samples,and use the system clustering analysis method to establish the identification model,and verify the classification results,the recognition accuracy reaches 91.67%;Using the identification model established by principal component analysis and support vector machine,the recognition accuracy can reach 98.75%.Fluorescence spectroscopy can be used to distinguish some samples by fluorescence detection.The accuracy of classification of fluorescence spectra of three kinds of bacteria can be 100%by principal component analysis. |
PDF Full Text Request