气相色谱—质谱联用仪的纯净谱图提取与检索算法的研究

发布时间：2018-01-28 02:20

本文关键词： 气相色谱-质谱联用仪定性分析色谱分辨率纯净质谱图提取质谱谱库检索　出处：《天津大学》2012年硕士论文　论文类型：学位论文

【摘要】：气相色谱-质谱联用仪(GC-MS)是对复杂混合物样品进行定性分析的重要工具。它融合了气相色谱和质谱两种仪器单独使用时的优点,具有更大的优势,广泛应用于环境、食品、地质检测等各个领域。原始GC-MS数据是GC-MS进行定性分析的基础,但它会包含来自仪器噪声、色谱柱流失物、基质等的干扰。同时,若混合物样品中包含保留时间接近的成分,得到的色谱图中会出现谱峰重叠的现象,一种成分的质谱图会包含来自其它物质离子的干扰。若直接应用原始GC-MS数据对混合物样品进行定性分析,产生的分析结果是不可靠的。因此,在定性分析之前,需要从原始的GC-MS数据中提取干净的谱图,以提高定性分析的准确性。最后,将得到的纯净质谱图与标准数据库中的谱图匹配,以完成定性分析。未知物质谱图与标准质谱图匹配采用计算机辅助的形式,其中,检索算法的实现是完成谱图匹配的关键。本论文以提高和改善纯净谱图提取算法与检索算法的性能为目标,以得到正确可靠的定性结果为目的,主要开展了两方面的工作。一是GC-MS纯净谱图提取算法的研究。本文首先将GC-MS数据表示成矩阵形式,并在此基础上提出了基于backfolding的纯净谱图提取算法。该算法旨在去除噪声的干扰,并从重叠的色谱峰中提取出独立的成分,最终得到纯净成分的质谱图,为下一步定性分析奠定基础。本论文的另外一个主要工作是质谱图检索算法的研究,提出了基于向量模型的检索算法,首先将未知物质谱图与标准质谱图表示成向量形式,通过比较向量间的相似程度来确定未知物质谱图与标准质谱图的相似性,最终实现可靠的定性分析。通过实验验证,采用本论文提出的纯净谱图提取算法与检索算法对原始GC-MS数据进行定性分析,得到的结果与现有商用软件的处理结果一致。本文提出的算法具有较好的性能,能够实现快速可靠的定性分析。
[Abstract]:Gas chromatography-mass spectrometry (GC-MS) is an important tool for qualitative analysis of complex mixture samples. It combines the advantages of gas chromatography and mass spectrometry when they are used alone and has more advantages. Widely used in the environment, food, geological detection and other fields. The original GC-MS data is the basis for qualitative analysis of GC-MS, but it will contain interference from instrument noise, chromatographic column loss, matrix, etc. At the same time. If the mixture contains components with close retention time, there will be overlapping peaks in the chromatogram obtained. The mass spectrogram of a component may contain interference from other material ions. If the original GC-MS data are directly used for qualitative analysis of the mixture sample, the analytical results are unreliable. In order to improve the accuracy of qualitative analysis, it is necessary to extract the clean spectrum from the original GC-MS data before qualitative analysis. Finally, the pure mass spectrum is matched with the spectrum in the standard database. In order to complete qualitative analysis, the matching of unknown matter spectrum and standard mass spectrum is in the form of computer aided, among which, the realization of retrieval algorithm is the key to complete spectrum matching. The aim of this thesis is to improve the performance of pure spectrum extraction algorithm and retrieval algorithm, and to obtain correct and reliable qualitative results. Two aspects of work are mainly carried out. One is the research of GC-MS pure spectrum extraction algorithm. Firstly, the GC-MS data is expressed in matrix form. On this basis, a pure spectrum extraction algorithm based on backfolding is proposed, which aims to remove noise interference and extract independent components from overlapping chromatographic peaks. Finally, the mass spectrogram of pure components is obtained, which lays a foundation for the next qualitative analysis. Another main work of this paper is the research of the mass spectrum retrieval algorithm, and a vector model-based retrieval algorithm is proposed. First, the unknown matter spectrum and the standard mass spectrum are expressed as vector forms, and the similarity between the unknown matter spectrum and the standard mass spectrum is determined by comparing the similarity between the vectors, and the reliable qualitative analysis is achieved. Through experimental verification, the pure spectrum extraction algorithm and retrieval algorithm proposed in this paper are used to qualitatively analyze the original GC-MS data. The proposed algorithm has good performance and can realize fast and reliable qualitative analysis.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.51

【参考文献】