中远红外光谱结合PLS-DA算法对非洲象牙与常见仿制品的鉴别  

Identification of African Ivory from Common Imitations by Mid-Far Infrared Spectroscopy Combined with PLSDA Algorithm

作　　者：王波 解迎双 王娟 张红艳 王慧珺 苏倡 蒲静 刘阿静 WANG Bo;XIE Ying-Shuang;WANG Juan;ZHANG Hong-Yan;WANG Hun-Jun;SU Chang;PU Jing;LIU A-Jing(Lanzhou Customs Technology Center,Lanzhou 730010;Northeast Forestry University,Harbin 150040;Science and Technology Research Center of China Customs,Beijing 100026)

机构地区：[1]兰州海关技术中心,兰州730010 [2]东北林业大学,哈尔滨150040 [3]中国海关科学技术研究中心,北京100026

出　　处：《中国口岸科学技术》2024年第S2期71-76,共6页China Port Science and Technology

基　　金：1+X合作单位资助;海关总署科研项目(2022HK008)

摘　　要：为了给打击象牙非法贸易提供技术支撑,本研究采集90份非洲象牙、牛骨和象牙果,利用中远红外光谱技术结合偏最小二乘判别分析(Partial least squares Discriminant Analysis,PLS-DA)算法对非洲象牙与市场上常见仿制品进行鉴别。通过筛选出的最佳预处理方法标准正态变量变换进行偏最小二乘回归建模,选取2个主成分因子和筛选15个特征波长点为模型的输入变量,结果表明建模集和预测集均为100%,中远红外光谱技术结合PLS-DA算法对非洲象牙的鉴别简便可靠且快速有效,同时为其他种类象牙制品的来源鉴别提供了新的思路。In order to provide technical support against the illegal trade in ivory,this study collected 90 samples of African ivory,bovine bones and Phytelephas macrocarpa,and used mid-far infrared spectroscopy technology combined with partial least squares discriminant analysis(PLS-DA)algorithm to identify African ivory from common imitations in the market.By employing the optimal preprocessing method,standard normal variable transformation for partial least squares regression(PLSR)modeling,two principal component factors and 15 selected characteristic wavelength points were used as input variables for the model.The results indicated that both the modeling set and the prediction set achieved 100%accuracy.The combination of mid-far infrared spectroscopy technology with the PLS-DA algorithm provides a simple,reliable,and rapid method for the identification of African elephant ivory.It also offers a novel approach for the source identification of other types of ivory products.

关 键 词：非洲象牙 中红外光谱 偏最小二乘判别分析 

分 类 号：O657.33[理学—分析化学] TS93[理学—化学]