基于电火花发射光谱的矿尘元素快速检测与分类算法研究  

Research on fast detection and classification algorithm of mineral dust elements based on spark emission spectroscopy

作　　者：何帅印 黄婧 李佳林 郑丽娜 HE Shuaiyin;HUANG Jing;LI Jialin;ZHENG Li’na(National Energy Shendong Coal Group Co.,Ltd.,Shenmu Shaanxi 719300,China;School of Safety Engineering,China University of Mining and Technology,Xuzhou Jiangsu 221000,China)

机构地区：[1]国家能源神东煤炭集团有限责任公司,陕西神木719300 [2]中国矿业大学安全工程学院,江苏徐州221000

出　　处：《中国安全生产科学技术》2025年第4期118-124,共7页Journal of Safety Science and Technology

基　　金：中国神华能源股份有限公司神东煤炭分公司科技创新项目(E21010028501);国家自然科学基金项目(52074274);中央高校基本科研业务费专项资金项目(2021YCPY0107)。

摘　　要：为解决矿尘成分检测技术耗时较长和设备昂贵问题,提出1种电火花发射光谱(SES)新技术,用于现场快速检测矿尘成分。通过SES设备,采集锰、铝、石灰石、硅和煤5种矿尘的光谱数据,分析各矿尘元素特征峰值,建立石灰石矿尘中Ca元素的校准曲线,研究SES法对矿尘的分类效果。研究结果表明:Mn, Al, Ca和Si元素的最强谱线分别位于344.2 nm, 309.27 nm, 317.93 nm和288.16 nm,而煤尘中Ca, Al, Mg, Si等元素的谱线也清晰可见。石灰石矿尘中Ca元素在442.43 nm处的校准曲线决定系数R^(2)为0.98,检测限为16 ppm。PCA分类结果显示,模型能以95%的方差解释率准确区分各类矿尘,外部验证中,锰、铝、煤尘的分类准确率达到100%。研究结果可为矿尘检测提供1种高效、低成本的新方法参考。To solve the problems of long time consumption and expensive equipment for detecting the mineral dust components,a new technology of spark emission spectroscopy(SES)for rapid on-site detection of mineral dust components was proposed.The spectral data of five types of mineral dust,namely manganese,aluminum,limestone,silicon,and coal,were collected using SES equipment.The characteristic peak values of each mineral dust element were analyzed,and a calibration curve for Ca element in limestone dust was established.The classification effect of SES method on mineral dust was studied.The results show that the strongest spectral lines of Mn,Al,Ca,and Si elements are located at 344.2 nm,309.27 nm,317.93 nm,and 288.16 nm,respectively,while the spectral lines of Ca,Al,Mg,Si,and other elements in coal dust are also clearly visible.The determination coefficient R^(2) of calibration curve for Ca element in limestone dust at 442.43 nm is 0.98,and the detection limit is 16 ppm.The PCA classification results show that the model can accurately distinguish various types of mineral dust with a variance explanation rate of 95%.In external validation,the classification accuracy of manganese,aluminum,and coal dust reaches 100%.The research results can provide an efficient and low-cost new method for the mineral dusts detection.

关 键 词：矿尘 电火花发射光谱 特征谱线 定量分析 

分 类 号：X964[环境科学与工程—安全科学]