搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

激光诱导击穿光谱技术结合神经网络和支持向量机算法的人参产地快速识别研究

董鹏凯 赵上勇 郑柯鑫 王蓟 高勋 郝作强 林景全

引用本文:
Citation:

激光诱导击穿光谱技术结合神经网络和支持向量机算法的人参产地快速识别研究

董鹏凯, 赵上勇, 郑柯鑫, 王蓟, 高勋, 郝作强, 林景全

Rapid identification of ginseng origin by laser induced breakdown spectroscopy combined with neural network and support vector machine algorithm

Dong Peng-Kai, Zhao Shang-Yong, Zheng Ke-Xin, Wang Ji, Gao Xun, Hao Zuo-Qiang, Lin Jing-Quan
PDF
HTML
导出引用
  • 利用激光诱导击穿光谱技术结合机器学习算法, 对东北5个产地(大兴安岭、集安、恒仁、石柱、抚松)的人参进行产地识别, 建立了主成分分析算法分别结合反向传播(BP)神经网络和支持向量机算法的人参产地识别模型. 实验采集了5个产地人参共657组在200—975 nm的激光诱导击穿光谱, 经光谱数据预处理后, 对C, Mg, Ca, Fe, H, N, O等元素的8条特征谱线进行主成分分析, 原光谱数据的前3个主成分累积贡献率达到92.50%, 且样品在主成分空间中呈现良好的聚集分类. 降维后的前3个主成分以2∶1进行随机抽取, 分别作为分类算法的训练集和测试集. 实验结果表明主成分分析结合BP神经网络及支持向量机的平均识别率分别为99.08%和99.5%. 发生误判的原因是集安和石柱两地地理环境的接近而导致的H, O两元素在Ca元素离子发射谱线下的归一化强度相似. 本研究为激光诱导击穿光谱技术在人参产地的快速识别提供了方法和参考.
    Based on laser-induced breakdown spectroscopy and machine learning algorithms, ginseng origin identification model is established by principal component analysis algorithm combined with back-propagation (BP) neural network and support vector machine algorithm to analyze and identify ginseng from five different origins in northeast China (Daxinganling, Ji’an, Hengren, Shizhu, and Fusong). The experiment collects a total of 657 groups of laser-induced breakdown spectral data from five origins of ginseng at 200–975 nm, reduces the background continuous spectrum of the original spectral data by moving window smoothing method, labels the ginseng LIBS spectral elements according to the American NIST atomic spectral database. Eight characteristic spectral lines of 7 elements Mg, Ca, Fe, C, H, N and O are selected for principal component analysis according to characteristic spectral selection conditions. The cumulative contribution rate of the first three principal components of the original spectral data reaches 92.50%, which represents a large amount of information about the original ginseng LIBS spectrum, and the samples show a good aggregation and classification in the principal component space. After dimension reduction, the first three principal components are randomly selected in a ratio of 2 to 1 and divided into 438 test sets and 219 training sets, which are used as the input values of the classification algorithm. The experimental results show that the principal component analysis combined with the BP neural network algorithm and support vector machine algorithm can correctly identify 217 and 218 spectra of 219 spectra of the test set respectively, and the average recognition rate is 99.08% and 99.5% respectively. The modeling time of BP neural network is 11.545 s shorter than that of the support vector machine. Both models misjudged Ji'an Ginseng as Shi zhu ginseng, and the reason for this misjudgment is that the normalized intensity of H and O under Ca element ion emission spectrum are similar due to the proximity of Ji 'an to Shi Zhu in geographical environment. The study presented here demonstrates that laser-induced breakdown spectroscopy combined with machine learning algorithm is a useful technology for rapid identification of ginseng origin and is expected to realize automatic, real-time, rapid and reliable discrimination.
      通信作者: 王蓟, jiji_w@163.com ; 高勋, gaoxun@cust.edu.cn
    • 基金项目: 国家自然科学基金(批准号: 61575030)、吉林省自然科学基金(批准号: 20180101283JC, 20200301042RQ, 20180201033GX, 20190302125GGX)和吉林省教育厅(批准号: JJKH20190539KJ)资助的课题.
      Corresponding author: Wang Ji, jiji_w@163.com ; Gao Xun, gaoxun@cust.edu.cn
    • Funds: Project supported by the National Natural Science Foundation of China (Grant No. 61575030), the Natural Science Foundation of Jilin province, China (Grant Nos. 20180101283JC, 20200301042RQ, 20180201033GX, 20190302125GGX), and the Research Foundation of Education Bureau of Jilin Province, China (Grant No. JJKH20190539KJ).
    [1]

    Patel S, Rauf A 2017 Biomed. Pharmacother. 85 120Google Scholar

    [2]

    Kim J H 2018 J. Ginseng Res. 42 264Google Scholar

    [3]

    Huang Y, Gou M J, Jiang K, Wang L J, Yin G, Wang J, Wang P, Tu J S, Wang T J 2019 Appl. Spectrosc. Rev. 54 653Google Scholar

    [4]

    Bec K B, Grabska J, Kirchler C G, Huck C W 2018 J. Mol. Liq. 268 895Google Scholar

    [5]

    Chen J B, Sun S Q, Ma F, Zhou Q 2014 Spectrochim. Acta, Part A 128 629Google Scholar

    [6]

    Fan Q, Chen C, Huang Z, Zhang C M, Liang P J, Zhao S L 2015 Spectrochim. Acta, Part A 136 1621Google Scholar

    [7]

    Radziemski L, Cremers D 2013 Spectrochim. Acta, Part B 87 3Google Scholar

    [8]

    邵妍, 张艳波, 高勋, 杜闯, 林景全 2013 光谱学与光谱分析 33 2593Google Scholar

    Shao Y, Zhang Y B, Gao X, Du C, Lin J Q 2013 Spectrosc. Spectr. Anal. 33 2593Google Scholar

    [9]

    Costa V C, Augusto A S, Castro J P, Machado R C, Andrade D F, Babos D V, Speranca M A, Gamela R R, Pereira E R 2019 Quim. Nova 42 527

    [10]

    Gottfried J L, Jr F C D L, Munson C A, Miziolek A W 2009 Anal. Bioanal. Chem. 395 283Google Scholar

    [11]

    Tzortzakis S, Anglos D, Gray D 2006 Opt. Lett. 31 1139Google Scholar

    [12]

    Kaiser J, Novotny K, Martin M Z, Hrdlicka A, Malina R, Hartl M, Adam V, Kizek R 2012 Surf. Sci. Rep. 67 233Google Scholar

    [13]

    谷艳红, 赵南京, 马明俊, 孟德硕, 贾尧, 方丽, 刘建国, 刘文清 2018 光谱学与光谱分析 38 982

    Gu Y H, Zhao N J, Ma M J, Meng D S, Jia Y, Fang L, Liu J G, Liu W Q 2018 Spectrosc. Spectr. Anal. 38 982

    [14]

    Choi J J, Choi S J, Yoh J J 2016 Appl. Spectrosc. 70 1411Google Scholar

    [15]

    Yao M Y, Yang H, Huang L, Chen T B, Rao G F, Liu M H 2017 Appl. Opt. 56 4070Google Scholar

    [16]

    Junjuri R, Gundawar M K 2019 J. Anal. At. Spectrom. 34 1683Google Scholar

    [17]

    Velioglu H M, Sezer B, Bilge G, Baytur S E, Boyaci Il H 2018 Meat Sci. 138 28Google Scholar

    [18]

    Lin J J, Lin X E, Guo L B, Guo Y M, Tang Y, Chu Y W, Tang S S, Che C J 2018 J. Anal. At. Spectrom. 33 1545Google Scholar

    [19]

    Wang J M, Liao X Y, Zheng P C, Xue S W, Peng R 2017 Anal. Lett. 51 575

    [20]

    郑培超, 郑爽, 王金梅, 廖香玉, 李晓娟, 彭锐 2020 光谱学与光谱分析 40 941

    Zheng P C, Zheng S, Wang J M, Liao X Y, Li X J, Peng R 2020 Spectrosc. Spectr. Anal. 40 941

    [21]

    Koujelev A, Sabsabi M, Ros V M, Laville S, Lui S L 2010 Planet. Space Sci. 58 682Google Scholar

    [22]

    于洋, 郝中骐, 李常茂, 郭连波, 李阔湖, 曾庆栋, 李祥友, 任昭, 曾晓雁 2013 物理学报 62 215201Google Scholar

    Yu Y, Hao Z Q, Li C M, Guo L B, Li K H, Zeng Q D, Li X Y, Ren Z, Zeng X Y 2013 Acta Phys. Sin. 62 215201Google Scholar

  • 图 1  激光诱导击穿光谱实验装置示意图

    Fig. 1.  Schematic diagram of the experimental setup of LIBS.

    图 2  BP神经网络结构示意图

    Fig. 2.  Structure of BP neural network.

    图 3  人参LIBS光谱(产地分别为大兴安岭、集安、恒仁、石柱、抚松)

    Fig. 3.  LIBS spectra of ginseng(the ginseng origins are DXAL, JA, HR, SZ and FS).

    图 4  (a)各主成分贡献率和主成分累积贡献率; (b)前3个主成分的三维散点图

    Fig. 4.  (a) Contribution rate of each principal component and cumulative contribution rate of principal component; (b) three-dimensional scatter plot of first three principal components.

    图 5  (a) BP神经网络训练性能曲线; (b) 分类结果图

    Fig. 5.  (a) BP neural network training performance curve; (b) classification results.

    图 6  (a) PCA-SVM网格参数优化; (b)分类识别结果图

    Fig. 6.  (a) PCA-SVM grid parameter optimization; (b) classification recognition result graph.

    图 7  人参LIBS谱中C, H, O元素谱线的归一化强度比

    Fig. 7.  Normalized intensity ratios of C, H and O element lines in the LIBS spectrum.

    表 1  人参特征谱线及波长

    Table 1.  Characteristic line and wavelength of ginseng.

    元素波长/nm
    C I247.80
    Mg II279.56
    Ca II393.40; 396.87
    Fe I422.71
    H I656.39
    N I747.07
    O I777.42
    下载: 导出CSV

    表 2  人参产地识别结果对比

    Table 2.  Comparison of ginseng origin identification results.

    算法测试集识别结果平均识别精度建模时间/s
    产地识别精度
    PCA-BP DXAL 100%
    JA 96%
    HR 100% 99.08% 2.48
    SZ 100%
    FS 100%
    PCA-SVM DXAL 100%
    JA 98%
    HR 100% 99.5% 14.03
    SZ 100%
    FS 100%
    下载: 导出CSV
  • [1]

    Patel S, Rauf A 2017 Biomed. Pharmacother. 85 120Google Scholar

    [2]

    Kim J H 2018 J. Ginseng Res. 42 264Google Scholar

    [3]

    Huang Y, Gou M J, Jiang K, Wang L J, Yin G, Wang J, Wang P, Tu J S, Wang T J 2019 Appl. Spectrosc. Rev. 54 653Google Scholar

    [4]

    Bec K B, Grabska J, Kirchler C G, Huck C W 2018 J. Mol. Liq. 268 895Google Scholar

    [5]

    Chen J B, Sun S Q, Ma F, Zhou Q 2014 Spectrochim. Acta, Part A 128 629Google Scholar

    [6]

    Fan Q, Chen C, Huang Z, Zhang C M, Liang P J, Zhao S L 2015 Spectrochim. Acta, Part A 136 1621Google Scholar

    [7]

    Radziemski L, Cremers D 2013 Spectrochim. Acta, Part B 87 3Google Scholar

    [8]

    邵妍, 张艳波, 高勋, 杜闯, 林景全 2013 光谱学与光谱分析 33 2593Google Scholar

    Shao Y, Zhang Y B, Gao X, Du C, Lin J Q 2013 Spectrosc. Spectr. Anal. 33 2593Google Scholar

    [9]

    Costa V C, Augusto A S, Castro J P, Machado R C, Andrade D F, Babos D V, Speranca M A, Gamela R R, Pereira E R 2019 Quim. Nova 42 527

    [10]

    Gottfried J L, Jr F C D L, Munson C A, Miziolek A W 2009 Anal. Bioanal. Chem. 395 283Google Scholar

    [11]

    Tzortzakis S, Anglos D, Gray D 2006 Opt. Lett. 31 1139Google Scholar

    [12]

    Kaiser J, Novotny K, Martin M Z, Hrdlicka A, Malina R, Hartl M, Adam V, Kizek R 2012 Surf. Sci. Rep. 67 233Google Scholar

    [13]

    谷艳红, 赵南京, 马明俊, 孟德硕, 贾尧, 方丽, 刘建国, 刘文清 2018 光谱学与光谱分析 38 982

    Gu Y H, Zhao N J, Ma M J, Meng D S, Jia Y, Fang L, Liu J G, Liu W Q 2018 Spectrosc. Spectr. Anal. 38 982

    [14]

    Choi J J, Choi S J, Yoh J J 2016 Appl. Spectrosc. 70 1411Google Scholar

    [15]

    Yao M Y, Yang H, Huang L, Chen T B, Rao G F, Liu M H 2017 Appl. Opt. 56 4070Google Scholar

    [16]

    Junjuri R, Gundawar M K 2019 J. Anal. At. Spectrom. 34 1683Google Scholar

    [17]

    Velioglu H M, Sezer B, Bilge G, Baytur S E, Boyaci Il H 2018 Meat Sci. 138 28Google Scholar

    [18]

    Lin J J, Lin X E, Guo L B, Guo Y M, Tang Y, Chu Y W, Tang S S, Che C J 2018 J. Anal. At. Spectrom. 33 1545Google Scholar

    [19]

    Wang J M, Liao X Y, Zheng P C, Xue S W, Peng R 2017 Anal. Lett. 51 575

    [20]

    郑培超, 郑爽, 王金梅, 廖香玉, 李晓娟, 彭锐 2020 光谱学与光谱分析 40 941

    Zheng P C, Zheng S, Wang J M, Liao X Y, Li X J, Peng R 2020 Spectrosc. Spectr. Anal. 40 941

    [21]

    Koujelev A, Sabsabi M, Ros V M, Laville S, Lui S L 2010 Planet. Space Sci. 58 682Google Scholar

    [22]

    于洋, 郝中骐, 李常茂, 郭连波, 李阔湖, 曾庆栋, 李祥友, 任昭, 曾晓雁 2013 物理学报 62 215201Google Scholar

    Yu Y, Hao Z Q, Li C M, Guo L B, Li K H, Zeng Q D, Li X Y, Ren Z, Zeng X Y 2013 Acta Phys. Sin. 62 215201Google Scholar

  • [1] 郭焱, 吕恒, 丁春玲, 袁晨智, 金锐博. 分数阶涡旋光衍射过程的机器学习识别. 物理学报, 2025, 74(1): 1-8. doi: 10.7498/aps.74.20241458
    [2] 侯佳佳, 张大成, 冯中琦, 朱江峰. 基于温度迭代校正自吸收效应的激光诱导击穿光谱定量分析方法. 物理学报, 2024, 73(5): 054205. doi: 10.7498/aps.73.20231541
    [3] 戴宇佳, 李明亮, 宋超, 高勋, 郝作强, 林景全. 空间约束结合梯度下降法提高铝合金中Fe成分激光诱导击穿光谱技术检测精度. 物理学报, 2021, 70(20): 205204. doi: 10.7498/aps.70.20210792
    [4] 刘立拓, 王春龙, 余晓娅, 石俊凯, 黎尧, 陈晓梅, 周维虎. 硅片表面纳米颗粒剥离及其成分检测方法研究. 物理学报, 2020, 69(16): 165201. doi: 10.7498/aps.69.20200517
    [5] 杨雪, 李苏宇, 姜远飞, 陈安民, 金明星. 不同样品温度下聚焦透镜到样品表面距离对激光诱导铜击穿光谱的影响. 物理学报, 2019, 68(6): 065201. doi: 10.7498/aps.68.20182198
    [6] 赵法刚, 张宇, 张雷, 尹王保, 董磊, 马维光, 肖连团, 贾锁堂. 基于自吸收量化的激光诱导等离子体表征方法. 物理学报, 2018, 67(16): 165201. doi: 10.7498/aps.67.20180374
    [7] 杨大鹏, 李苏宇, 姜远飞, 陈安民, 金明星. 飞秒激光成丝诱导Cu等离子体的温度和电子密度. 物理学报, 2017, 66(11): 115201. doi: 10.7498/aps.66.115201
    [8] 杨文斌, 周江宁, 李斌成, 邢廷文. 激光诱导氮气等离子体时间分辨光谱研究及温度和电子密度测量. 物理学报, 2017, 66(9): 095201. doi: 10.7498/aps.66.095201
    [9] 刘玉峰, 张连水, 和万霖, 黄宇, 杜艳君, 蓝丽娟, 丁艳军, 彭志敏. 激光诱导击穿火焰等离子体光谱研究. 物理学报, 2015, 64(4): 045202. doi: 10.7498/aps.64.045202
    [10] 刘玉峰, 丁艳军, 彭志敏, 黄宇, 杜艳君. 激光诱导击穿空气等离子体时间分辨特性的光谱研究. 物理学报, 2014, 63(20): 205205. doi: 10.7498/aps.63.205205
    [11] 张颖, 张大成, 马新文, 潘冬, 赵冬梅. 基于激光诱导击穿光谱技术定量分析食用明胶中的铬元素. 物理学报, 2014, 63(14): 145202. doi: 10.7498/aps.63.145202
    [12] 陈添兵, 姚明印, 刘木华, 林永增, 黎文兵, 郑美兰, 周华茂. 基于多元定标法的脐橙Pb元素激光诱导击穿光谱定量分析. 物理学报, 2014, 63(10): 104213. doi: 10.7498/aps.63.104213
    [13] 郭连波, 郝荣飞, 郝中骐, 李阔湖, 沈萌, 任昭, 李祥友, 曾晓雁. 激光诱导AlO自由基B2+X2+跃迁光谱研究. 物理学报, 2013, 62(22): 224211. doi: 10.7498/aps.62.224211
    [14] 张旭, 姚明印, 刘木华. 激光诱导击穿光谱结合偏最小二乘法定量分析脐橙中Cd含量. 物理学报, 2013, 62(4): 044211. doi: 10.7498/aps.62.044211
    [15] 王春龙, 刘建国, 赵南京, 马明俊, 王寅, 胡丽, 张大海, 余洋, 孟德硕, 章炜, 刘晶, 张玉钧, 刘文清. 水体重金属激光诱导击穿光谱定量分析方法对比研究. 物理学报, 2013, 62(12): 125201. doi: 10.7498/aps.62.125201
    [16] 于洋, 郝中骐, 李常茂, 郭连波, 李阔湖, 曾庆栋, 李祥友, 任昭, 曾晓雁. 支持向量机算法在激光诱导击穿光谱技术塑料识别中的应用研究. 物理学报, 2013, 62(21): 215201. doi: 10.7498/aps.62.215201
    [17] 鲁翠萍, 刘文清, 赵南京, 刘立拓, 陈东, 张玉钧, 刘建国. 土壤重金属铬元素的激光诱导击穿光谱定量分析研究. 物理学报, 2011, 60(4): 045206. doi: 10.7498/aps.60.045206
    [18] 孙对兄, 苏茂根, 董晨钟, 王向丽, 张大成, 马新文. 基于激光诱导击穿光谱技术的铝合金成分定量分析. 物理学报, 2010, 59(7): 4571-4576. doi: 10.7498/aps.59.4571
    [19] 张大成, 马新文, 朱小龙, 李 斌, 祖凯玲. 激光诱导击穿光谱应用于三种水果样品微量元素的分析. 物理学报, 2008, 57(10): 6348-6353. doi: 10.7498/aps.57.6348
    [20] 李宏斌, 刘文清, 张玉钧, 丁志群, 赵南京, 魏庆农, 王玉平, 杨立书. 基于径向基函数网络的激光诱导荧光特征光谱分离算法. 物理学报, 2005, 54(9): 4451-4457. doi: 10.7498/aps.54.4451
计量
  • 文章访问数:  7832
  • PDF下载量:  174
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-09-11
  • 修回日期:  2020-10-19
  • 上网日期:  2021-02-06
  • 刊出日期:  2021-02-20

/

返回文章
返回