激光诱导击穿光谱矿浆品位分析仪在铁矿选矿过程中的应用
Application of laser-induced breakdown spectroscopy slurry grade analyzer in iron ore beneficiation
胡健 1盖俊鹏 1齐立峰 2王金池 2郑黎明 2孙兰香3
作者信息
- 1. 鞍钢集团关宝山矿业有限公司,辽宁鞍山 114044
- 2. 中国科学院沈阳自动化研究所机器人学国家重点实验室,辽宁沈阳 110016
- 3. 中国科学院沈阳自动化研究所机器人学国家重点实验室,辽宁沈阳 110016;辽宁辽河实验室,辽宁沈阳 110169
- 折叠
摘要
基于激光诱导击穿光谱(LIBS)技术的矿浆品位分析仪,通过管道取样器自动取样,将矿浆样本直接输送至分析仪器,实现选矿生产过程中矿浆品位在线测量.针对铁矿选矿生产过程中铁矿矿浆的特点,对二次缩分取样量及激光能量和光斑大小等参数进行了优化分析,确定二次缩分取样量为10 L,激光能量为200 mJ,聚焦光斑大小为1.0 mm.结合基于循环变量筛选的非线性偏最小二乘法进行了模型建立,并结合不同阶段的同步取样分析对比,结果显示矿浆品位分析仪在测量矿浆Fe品位时,其平均绝对误差(MAE)与均方根误差(RMSE)均低于1％,能够满足生产控制需求.该技术突破了铁矿浆铁品位在线精准测量的瓶颈,有力推动了矿山智能化生产的发展,为构建智能工厂奠定了重要的基础.
Abstract
A slurry grade analyzer based on laser-induced breakdown spectroscopy technology was employed for the online measurement of slurry grade in the production process of iron ore beneficiation.The automat-ic sampling was conducted through a pipeline sampler,and the slurry sample was directly transported to the analyzer.Based on the characteristics of iron ore slurry in the production process of iron ore beneficiation,the parameters of analyzer system such as secondary fractional sampling volume,laser energy,and spot size were optimized.The optimal secondary fractional sampling volume was 10 L,the laser energy was 200 mJ,and the spot size was 1.0 mm.A model was established using the nonlinear partial least squares(PLS)method based on cyclic variable filtering,and then compared with synchronous sampling analysis at differ-ent stages.The results showed that the mean absolute error(MAE)and root mean square error(RMSE)were both less than 1％when the Fe grade in slurry was measured by the slurry grade analyzer,which could meet the requirements of production control.This technology had broken the bottleneck of precise online measurement of iron grade in slurry.It had effectively promoted the development of intelligent min-ing production and laid an important foundation for building an intelligent factory.
关键词
选矿/激光诱导击穿光谱(LIBS)/铁矿浆/品位/在线分析/偏最小二乘法Key words
beneficiation/laser-induced breakdown spectroscopy(LIBS)/iron ore slurry/grade/online analy-sis/partial least squares(PLS)引用本文复制引用
基金项目
国家自然科学基金(62173321)
辽宁辽河实验室项目(LLL23ZZ-05-02)
出版年
2024
NETL
NSTL
万方数据