首页 | 本学科首页   官方微博 | 高级检索  
     检索      

饲料添加剂L-赖氨酸硫酸盐中L-赖氨酸含量近红外速测方法研究
引用本文:李守学,陈玉艳,贾铮,肖志明,王燕妮,刘成新,樊霞.饲料添加剂L-赖氨酸硫酸盐中L-赖氨酸含量近红外速测方法研究[J].动物营养学报,2017,29(10).
作者姓名:李守学  陈玉艳  贾铮  肖志明  王燕妮  刘成新  樊霞
作者单位:1. 中国农业科学院农业质量标准与检测技术研究所,北京 100081;中国农业大学工学院,北京 100083;2. 辽宁省兽药饲料畜产品质量安全检测中心,沈阳,110016;3. 中国农业科学院农业质量标准与检测技术研究所,北京,100081
基金项目:"十二五"农村领域国家科技计划课题,中国农业科学院"饲料质量安全检测与评价"创新团队项目资助
摘    要:为了探讨利用近红外漫反射光谱技术(NIDRS)快速定量分析饲料添加剂L-赖氨酸硫酸盐中L-赖氨酸含量的可行性,本试验在全国范围内收集了具有代表性的L-赖氨酸硫酸盐添加剂76个,采用国家标准方法对样品中的L-赖氨酸含量进行化学赋值;用光栅型近红外光谱仪扫描L-赖氨酸硫酸盐样品,获取了不同物理状态下样品的近红外光谱图。依据L-赖氨酸含量将样品分为定标集和验证集,运用适当的光谱预处理方法,采用竞争性自适应重加权(CARS)算法结合偏最小二乘法(PLS)建立了L-赖氨酸硫酸盐的近红外定标分析模型,并将该模型与全波长模型进行了比较。结果表明:用烘干、60目粉碎后的样品结合CARS算法建立的定标模型最优,定标集校正决定系数(R2C)为0.954,校正集标准偏差(SEC)为0.510,交互验证标准偏差(SECV)为0.659;验证集预测决定系数(R2P)为0.952,预测标准偏差(SEP)为0.554,相对分析误差(RPD)值为3.83。由此可见,NIDRS定量分析L-赖氨酸硫酸盐具有一定可行性,对于丰富我国氨基酸盐及其他氨基酸制品的快速检测方法具有实际的应用意义。

关 键 词:赖氨酸  近红外漫反射光谱技术  速测模型  竞争性自适应重加权算法

Near Infrared Rapid Determination Method for L?Lysine Content in Feed Additive L?Lysine Sulfate
LI Shouxue,CHEN Yuyan,JIA Zheng,XIAO Zhiming,WANG Yanni,LIU Chenxin,FAN Xia.Near Infrared Rapid Determination Method for L?Lysine Content in Feed Additive L?Lysine Sulfate[J].Acta Zoonutrimenta Sinica,2017,29(10).
Authors:LI Shouxue  CHEN Yuyan  JIA Zheng  XIAO Zhiming  WANG Yanni  LIU Chenxin  FAN Xia
Abstract:In order to explore the feasibility of near infra diffuse reflectance spectroscopy ( NIDRS) method for rapidly quantitative determination of L?lysine content in feed additive L?lysine sulfate, seventy?six representative samples of L?lysine sulfate were collected in China. According to the national standard method to chemical eval?uated the L?lysine content of samples; the near infrared spectrogram of L?lysine sulfate samples at different physical states were scanned by raster near infrared diffuse reflectance spectroscopy instrument. On the basis of L?lysine content samples were divided into calibration and validation sets, using appropriate spectral pretreat?ment method, the near infrared quantitative analysis model of L?lysine sulfate was established by using competi?tive adaptive reweighted sampling ( CARS) algorithm combined partial least square method ( PLS) , and com?pared with the full wavelength model. The results showed that the optimal calibration model was established with the sample of dry and crushed at 60 mesh, and the determination coefficient of correction calibration set (R2C) was 0.954, the standard error of calibration (SEC) was 0.510, the standard error of cross validation (SECV) was 0.659. The determination coefficient of prediction validation set (R2P) was 0.952, the standard error of prediction ( SEP) was 0.554, the relative percent deviation ( RPD) value was 3.83. In conclusion, NIDRS quantitative analysis of L?lysine sulfate has certain feasibility, it has practical application significance for rich other amino acids fast detection methods in our country.
Keywords:lysine  NIDRS  rapid determination model  CARS
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号