基于多元校正法的香梨糖度可见近红外光谱检测.PDFVIP

20 10 年 12 月 农 业 机 械 学 报 第41卷 第 12 期 DOI:10.3969/ j.issn.1000鄄1298.2010.12.026 基于多元校正法的香梨糖度可见/ 近红外光谱检测* 徐惠荣摇 陈晓伟摇 应义斌 (浙江大学生物系统工程与食品科学学院,杭州310029) 摇 摇 揖摘要铱摇 应用可见/ 近红外光谱透射技术结合多元校正法探讨了样品在不同温度条件(5、10、15、20益)下香梨 糖度的快速无损检测。 在波长500~900nm范围内,用逐步多元线性回归(SMLR)、偏最小二乘法(PLS)、最小二乘 支持向量机(LS SVM1、LS SVM2)和遗传算法 偏最小二乘法(GA PLS)等多种多元校正法进行了建模预测比 较研究。 预测结果从优到差依次为LS SVM2、LS SVM1、GA PLS、PLS、SMLR。 关键词:香梨摇 糖度摇 无损检测摇 可见/ 近红外光谱摇 多元校正摇 温度 中图分类号:S123 文献标识码:A 文章编号:1000鄄1298(2010)12鄄0126鄄04 Multivariate Approach to Determinate Sugar Content of Fragrant Pears with Temperature Variation by Visible/ NIR Spectroscopy Xu Huirong摇 Chen Xiaowei摇 Ying Yibin (School of Biosystems Engineering and Food Science,Zhejiang University,Hangzhou310029,China) Abstract Nondestructive determination of sugar content (SC) of fragrant pears with temperature variation (5益,10益,15益 and 20益)was studied by visible鄄NIR spectroscopy coupled with multivariate calibration methods. In the region of 500~900 nm, five multivariate calibration models: stepwise multiple linear regression (SMLR),partial least squares (PLS),least squares鄄support vector machines (LS SVM1, only 30 bands spectral data obtained by SMLR analysis inputted as X variable), LS SVM2 (both 30 bands spectral data and sample temperature inputted as X variable) and genetic algorithm鄄partial least squares (GA PLS) were compared. The results showed that the prediction performance was in the order of LS SVM2,LS SVM1,GA PLS,PLS and SMLR. Key words 摇 Fragrant pears, Sugar content, Nondestructive determination, Visible鄄NIR spectroscopy,Multivariate calibration,Temperature