基于近红外光谱技术结合遗传算法的安溪铁观音品质评价  

The quality evaluation of Anxi Tieguanyin tea based on near-infrared spectroscopy combined with genetic algorithm

作　　者：夏金梅[1,2] 孙威江[2] 林涛[1] 游泳[1] 曾兆华[1] 王冰玉[2] 

机构地区：[1]福建省农业科学院植物保护研究所农药环境安全评价中心,福建福州350002 [2]福建农林大学园艺学院,福建福州350002

出　　处：《武夷科学》2016年第1期132-137,共6页Wuyi Science Journal

基　　金：福建省自然科学基金(2016J01139);福建省省属公益类科研院所专项(2014R1024-6;2016R1023-9)

摘　　要：以清香型铁观音茶叶为试验原料,基于近红外光谱技术结合遗传算法,建立相关的数学分析模型,用于安溪铁观音综合品质得分的测定,完善茶叶检测体系,进一步推动我国茶产业的标准化进程。实验结果表明,经二阶导数+平滑+归一化方法对光谱进行预处理后,得到综合品质得分PLS测定模型最优,验证集相关系数为0.913,均方根偏差为0.665。选用近红外光谱6 670-4 000 cm-1谱区,经遗传算法筛选特征波长后,建立茶样综合品质得分GA-PLS测定模型,校正集相关系数为0.959,均方根偏差为0.413;验证集相关系数为0.940,均方根偏差为0.587,GA-PLS模型的预测能力和精度更高。To evaluate the quality of Anxi Tieguanyin （AXTGY） tea by unified standards, the present experiment studied the characterization of AXTGY tea using near-infrared spectroscopy and genetic algorithm. AXTGY tea was collected as the research materials, and genetic algorithm （GA） combined with partial least squares （PLS） was used for building the mathematical analysis model to analyse the quality of AXTGY tea. The results of comprehensive quality score determination model showed that the PLS model displayed the highest prediction performance after spectrum being processed by smoothing, the second derivative and normalized methods. Statistic results of PLS model： Rp = 0.913, RMSEP = 0.665.NIR spectra ranged from 6 670 to 4 000 cm-1 were selected, using GA to select characteristic wavelength and then to construct GA-PLS calibration model which has higher predictive power and precision. Statistic results of GA-PLS model were Rc = 0.959, RMSEC = 0.413, Rp = 0.940, RMSEP = 0.587.

关 键 词：近红外光谱技术 遗传算法 偏最小二乘法 安溪铁观音 品质评价 

分 类 号：O657.33[理学—分析化学] S132[理学—化学]