基于激光诱导击穿光谱煤质分析仪的研制

发布时间：2018-01-09 19:15

本文关键词：基于激光诱导击穿光谱煤质分析仪的研制　出处：《山西大学》2015年硕士论文　论文类型：学位论文

【摘要】：激光诱导击穿光谱(Laser-Induced Breakdown Spectroscopy, LIBS)是一种新兴的激光烧蚀光谱分析技术。该技术基于原子光谱与离子光谱的波长与特定的元素一一对应,光谱强度与对应元素的含量具有一定的量化关系的基本原理,利用高功率脉冲激光与物质相互作用产生等离子体,采用光学系统对等离子体中的原子和离子发射光谱进行收集,根据对应的光谱实现对物质化学元素的定性与定量分析。该技术由于具有无需样品预处理,可多元素同时分析,速度快,无辐射等优势,已经逐渐被应用到各行各业中。例如,在快速或在线煤质检测方面。目前国内外一些单位已经开展了LIBS应用于燃煤电厂快速煤质检测方面的研究,并取得了一定的成果。但大多数还主要停留在实验室阶段,尚未真正实现工业化应用和推广。本文首先介绍了煤质检测的现状,LIBS技术的先进性、基本原理、定量分析原理及技术特点；然后搭建了LIBS煤质分析实验装置,系统的优化了LIBS实验装置参数,发展了脉冲激光功率稳定系统,使得脉冲激光功率长时间运行时始终位于预设的区间内,保证了长期测量的稳定性；发展了合理的光谱预处理方法,并采用支持向量机和偏最小二乘法建立了煤质预测模型；在以上研究的基础上,开发了可快速检测煤样中灰分、挥发分、发热量等工业指标的LIBS煤质分析仪,并进行了工业现场测试,推动了该技术的实用化。论文就以上内容分为以下四个部分：1.介绍了激光诱导击穿光谱技术的应用背景和现状,详细的了解了等离子体形成过程、等离子体元素谱线产生的过程和展宽机制、LIBS技术的特点。2.建立了LIBS煤质分析实验装置,对装置中的关键参数进行了系统的优化,针对该技术在实用化当中,激光器输出的功率会由于长时间测量产生慢漂移的问题,我们建立了闭环负反馈式激光功率稳定系统,使得激光功率能够稳定在设定的区间,从而保证了长时间测量的稳定性。3.构建了实用有效的LIBS光谱数据预处理方法,包括采用洛伦兹与线型相结合的方式对光谱进行拟合并进行归一化与筛选。采用支持向量机和偏最小二乘法建立煤炭灰分工业分析指标的定标模型,利用上述两种模型对煤灰分值进行预测,预测的均方根误差分别为4.4%和4.5%。4.研制了适用于化验室分析的光电一体化的全自动激光煤质分析仪,可在120s内完成对煤样的灰分、挥发分、发热量等工业指标的测量。
[Abstract]:Laser induced breakdown spectroscopy (Laser-Induced Breakdown Spectroscopy, LIBS) is a new kind of laser ablation spectroscopy. The technique is based on atomic spectra and ion spectral wavelength and specific elements corresponding to the basic principle and content of spectral intensity corresponding to the elements with a certain quantitative relationship, using high power pulsed laser interaction with matter. The optical system to produce plasma, plasma atomic and ionic emission spectra were collected, according to the qualitative and quantitative analysis of chemical elements of the corresponding spectrum. This technique has no sample pretreatment, simultaneous multielement analysis, fast speed, no radiation and other advantages, has been gradually applied to all walks of life. For example, in the fast or online coal quality testing at home and abroad. Some units have carried out the application of LIBS in the coal-fired power plant soon Study on fast coal quality detection, and achieved certain results. But most are still in the laboratory stage, has not yet realize the application and promotion of industrialization. This paper introduces the present situation of coal quality testing, the basic principle of the advanced LIBS technology, quantitative analysis, principle and technical features; then experiments of LIBS coal quality analysis the optimization of the system parameters, LIBS experimental device, the development of pulse power system stability, the operation interval of pulse laser power long time always located in the default, to ensure a stable long-term development of measurement; spectrum pretreatment method is reasonable, and the establishment of the coal quality prediction model and partial least squares support vector machine; on the basis of the above research, the development of the rapid detection of the coal ash, volatile matter, LIBS coal analyzer calorific industrial index, and the The industrial field test, promote the practical technology. The above content is divided into the following four parts: 1. introduce the application background and the present situation of laser induced breakdown spectroscopy, detailed understanding of the plasma formation process, plasma spectral lines broadening mechanism of the process and the characteristics of LIBS technology,.2. established experiment LIBS device for coal quality analysis, optimization of the key parameters in the device, according to the technology in the practical, power output due to long time measurement of slow drift problem, we established a closed-loop negative feedback laser power stabilization system, the laser power can be stabilized in the set range, to ensure.3. long time stability measurement method to construct LIBS spectral data preprocessing is practical and effective, including the use of Lorenz and linear combination of spectra For fitting and normalization and screening. The calibration model was established index analysis of coal ash industrial support vector machine and partial least squares method to predict the ash score using the above two models, the root mean square error of prediction for photoelectric integration 4.4% and 4.5%.4. was developed for laboratory analysis respectively full automatic laser coal analyzer. Can complete the ash of coal samples in 120s volatile, measuring the calorific value and other industrial index.

【学位授予单位】：山西大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN249;O433

【参考文献】