Design The Micro-Raman Spectroscopy And Study On Data Processing Of Raman Spectrum

Raman spectra can reflect the sample characteristics of some specific elements,Therefore, a qualitative analysis of Raman spectra can distinguish between similar objects through a qualitative analysis and quantitative analysis of the function of the concentration. Raman spectra analysis generally do not touch and need not take any modification of the sample. It can pass through the glass or precious stones container to collect Raman information. In the industrial production, do not need for sample pre-treatment is one of the main reason to use Raman spectroscopy. Low maintenance costs, specific analysis capabilities, Raman spectroscopy and infrared spectroscopy complementary characteristics ,So Raman spectroscopy has became one of research hotspot of analytical science field. This thesis designed a Raman spectrometer and researched the data disposal of Raman spectroscopy. The main contributions of this thesis are as follows:introduce the research background and significance of Raman spectroscopy, explain the search status at home and abroad of Raman spectra Instrument;Explain the Raman scattering respectively from the perspective of classical theory and quantum theory. Expounded the basic principles and the basic classification structure of Raman spectroscopy instrument.;According to micro-Raman Spectrometer design goals, design the holographic notch filter meet the use requirements of the holographic notch filter and draw the filter transmittance curve. Calculate the basic parameters of grating determine the spectral parameters of chern-turner spectrophotometric system. Detailed describe the sample optical, Identified incident optical way, Sample holder, scattered light collection. According to the design requirements, Select the appropriate detector and laser.;Introduced Fourier transform filtering and Wavelet threshold filtering, the above methods can not fundamentally resolve the spectral fluorescence background interference. Due to the defects of Donoho's method in de-noising of Laman spectral signal, a new thresholding function was presented based on the wavelet threshold function put forward by D.L.Donoho and I.M.Johnstone. It overcomes an invariable dispersion of the soft-thresholding method between the estimated wavelet coefficients and the decomposed wavelet coefficients, and eliminate the oscillation defect of the hard-threshold denoising method. At the same time,the new thresholding function is continuous as the soft-thresholding function and more elastic than the hard-thresholding function or soft-thresholding function. The result shows that the new method can gave better SNR and MSE gains than hard-thresholding or soft-thresholding methods. ; When carrying out quantitative analysis, use PLS-SVM method, model the methanol. The results show that PLS-SVM is better than PLS in predicted results.