Research Of System Design Of Raman System And Raman Spectroscopy Data Analysis Based On Genetic Algorithm

Raman spectroscopy (RS) is a kind of molecule scattering spectroscopy, which is characterized by the frequency excursion that caused by interactions of molecule and photon to show the information of molecule. The frequency excursion reflects the margin between some certain vibrational or rotational energy states. Every substance has its own feature Raman spectroscopy, which intensity is proportional to its concentration under given condition with other factors fixed. This is the basis of Raman spectroscopy to analyze the structure, components, concentrations and some other properties of samples. There are many advantages to investigate matter with RS, such as simple system structure with relative lower cost, wide variety of detectable mass, nondestructive detect for multicomponent sample, good sensitivity, feasibility of real-time analysis and online examination assisted with optical fiber and computers, etc. The key to realize the qualitative and quantitative analysis is accurate Raman spectroscopy detecting system and high efficient computer technique for system control, data process and analysis. It's still an open question for thorough discussion to the analysis of mixture's Raman spectroscopy.In this thesis, we first introduce the main theory of Raman spectroscopy, including the mechanism it occurs, the physical explanation, the basis of quantitative analysis, etc. Then new types of Raman technique such as NIR-FT Raman, RRS, SERS, microscope Raman and co-focus Raman, fiber Raman, etc. are discussed. And the main applications of Raman spectroscopy on chemistry and material science, medicine, biology, judging and identifying science, industry, environmental protection, geology and archeology, etc. are introduced. In the 3rd Chapter, on the basis of modern Raman instrument, we use separate elements to build a Raman system. And the calibration and realization of important parameters of one of the main part, CCD detector, is discussed. In the 4th Chapter, we first introduce basic concept, compute flow, main elements, and trait of the GeneticAlgorithm. Then we use our G-A program to analyze the Raman spectroscopy data of mixed sample composed of ten species of ore. We run the G-A program repeatedly under certain condition and then do the experiment under different conditions. The affection of diversity of algorithm and difference of its parameters on the performance is discussed. From the result, it can be argued that the G-A program is of good feasibility, high efficiency and nice accuracy in quantitative analysis of mixture's Raman spectroscopy. This shows the potential capability of Genetic Algorithm in the application of quantitative analysis of mixture's Raman spectroscopy. Finally, the present stage study and work are summarized and some future work is discussed.