小型化远程激光诱导击穿光谱仪器研发及对合金钢中硅钼铬的分析
Development of miniaturized remote laser-induced breakdown spectroscopy instrument and its application in analysis of silicon,molybdenum and chromium in alloy steel
宋骁冉 1牛广辉 2汤悦 3许应铜 3林庆宇 3段忆翔3
作者信息
- 1. 桂林理工大学机械与控制工程学院,广西桂林 541004
- 2. 宝武重工有限公司研发中心,上海 201900
- 3. 四川大学机械工程学院,四川成都 610064
- 折叠
摘要
为提升激光诱导击穿光谱(LIBS)技术在各种极端环境下的灵活分析能力,本研究实现了远程LIBS仪器的小型化设计.仪器采用模块化设计,以云台为底座,能够实现对远距离目标的灵活检测分析,采用8种合金钢标准样品为实验对象,在2 m条件下进行光谱采集,对标样中Si、Mo及Cr这3种元素进行定量分析.使用随机森林算法(RF)分别构建了上述3种元素的定量模型,模型决定系数分别为0.999 8、0.998 8、0.999 1.在对模型验证时,验证样品中Si、M o及Cr 3种元素预测含量的相对误差最小值分别为14.6％、12.9％、11.6％.所研发的小型化远程LIBS仪器具有良好的远程定量检测能力.
Abstract
In order to improve the flexible analysis capability of laser-induced breakdown spectroscopy(LIBS)under various extreme environments,the miniaturization design of remote LIBS instrument was re-alized in this study.The instrument adopted modular design,and the tripod head was used as base sup-port.It was applicable for the flexible detection and analysis of remote targets.8 certified reference mate-rials of alloy steel were used as the experimental subject.The spectral acquisition was carried out under the condition of 2 m.3 elements in certified reference materials,including Si,Mo and Cr,were quantitatively analyzed.The quantitative models of 3 elements were established by random forest(RF)algorithm,and the coefficients of determination were 0.999 8,0.998 8 and 0.999 1,respectively.During the model vali-dation,the minimum relative errors of predicted contents of 3 elements(Si,Mo and Cr)in validation sam-ples were 14.6％,12.9％and 11.6％,respectively.The developed miniaturized remote laser-induced breakdown spectroscopy instrument showed good remote quantitative detection capability.
关键词
激光诱导击穿光谱(LIBS)/小型化/远程分析/合金钢/硅/钼/铬/随机森林(RF)Key words
laser-induced breakdown spectroscopy(LIBS)/miniaturized/remote analysis/alloy steel/silicon/molybdenum/chromium/random forest(RF)引用本文复制引用
出版年
2024
NETL
NSTL
万方数据