Establishment And Application Of Laser Induced Breakdown Spectroscopy Database And Energetic Material Database

Database has received researchers' widespread attentions due to its large capacity, efficiency and reliability, it puts forward higher requirements and urgent needs to establish chemistry database for the rapid development and industrial demands of chemical disciplines. Wherein, with the rapid development of laser-induced breakdown spectroscopy(LIBS) technology and energetic materials, it is urgently necessary to establish a series of databases on LIBS, energetic materials and its synthetic routes. In this thesis, based on some practical problems on the online and in-site analysis of LIBS technique, screening and synthesis of high-performance of energetic materials, it combined with computer, database technology and chemometrics methods. At first, the basic and metallurgical exclusive LIBS databases were established for the identification of steel grades and slag types, and constructed the preprocessing and analysis system on LIBS spectra. Moreover, the base compound database, reaction databases and spectral database of energetic materials were established and applied for screening, virtual synthesis and spectral matching analysis of energetic materials, after which the relevant software systems were built. The present studies not only enriched chemometrics and chemical databases, and further laid the foundation for developing in-situ analysis application based on LIBS, but also provides new methods for the development of high-performance energetic materials, and further provide information, methodologies and technical support for the research and production of metallurgy, geology, oil and related industries. There are four chapters for this thesis and the main contents are as follows:1. Based on LIBS spectra, the basic and metallurgical exclusive LIBS databases were established and were applied for the identification of steel grades and slag types. For the steel grade identification, the element comparison method and similarity matching method based on LIBS database were proposed; the variable importance random forest (VIRF) method based on LIBS database was proposed to identify slag type. On the basis of constructed modeling methods, we constructed the preprocessing and analysis system on LIBS spectra. Compared with the traditional discriminant analysis method on steel grades and slag types, the three proposed methods shows some advantages of without complex sample preparation, high accuracy, fast and simple analysis; the built preprocessing and analysis system on LIBS spectra can not only fast processing for LIBS data, and also can be used for automatic online identification and qualitative analysis of metallurgy, geology, oil and other fields.2. Based on energetic materials, the base compound database, reaction knowledge databases and spectral database of energetic materials were established and were applied for screening, virtual synthesis and spectral matching analysis of energetic materials. For the screening of energetic materials, the neural network partial least squares (NNPLS) and genetic algorithm support vector machine (GASVM) based on the base compound database were proposed to predict the detonation properties of energetic materials, and the performance estimations system of energetic materials was constructed based on the two modeling and related methods; For the virtual synthesis of energetic compounds, the encoding methods, matching algorithms and split rules of chemical molecular structure were investigated, and then designed and constructed the virtual synthesis system of energetic compounds. For the spectra database management, the fast spectral matching algorithm based on the spectra database was studied according to four categories (IR, MS and so on) spectra features, and constructed the database management system on energetic compounds spectra. It shows that the three constructed database systems can be applied for the detonation properties calculations, virtual synthesis and qualitative analysis of energetic compound, and shows well performances such as simple operation, friendly interface and reliable, and further shows the characteristics of improving screening hit rate and shortening the development period of energetic compound. The thesis has enriched chemometrics and chemical databases, provided new methods for the development of high-performance energetic materials, and further provided information, methodologies and technical support for the research of energetic materials, functional materials and drug.