【摘要】 油菜机械化收获的油菜籽叶绿素含量不一致,直接影响其加工成本和菜籽油品质,亟需建立成熟期油菜籽中叶绿素快速无损检测技术。本文采用紫外可见分光光度法测定450份代表性油菜籽样品中叶绿素含量,并采集近红外光谱数据。通过一阶导数、标准正态变量变换预处理,采用竞争性自适应重加权采样算法进行波长选择,建立了油菜籽中叶绿素含量偏最小二乘法回归模型。交互检验结果显示,油菜籽中叶绿素含量预测模型的验证集决定系数为0. 944 6,交叉检验均方根误差(RMSECV)为1. 36mg/kg。研究表明,该方法可以准确预测油菜籽中叶绿素含量,为油菜籽品质快速监测提供了重要的技术支撑。更多还原

【Abstract】 With the advancement of mechanization of rapeseed production in China,chlorophyll content of rapeseed increases significantly due to the inconsistent maturity of silique,which directly affects the quality of rapeseed oil and increases the processing cost. Therefore,establishment of rapid and non-destructive detection techniques for chlorophyll in rapeseed is important to quality control of rapeseed oil. In this study,the chlorophyll contents of 450 representative rapeseeds were quantitatively analyzed by UV-Vis spectrophotometry. Subsequently,these rapeseeds were analyzed by near infrared spectroscopy( NIR). After the pretreatment of the first derivative and standard normal variate( SNV),and wavelength selection by competitive adaptive reweighted sampling( CARS),partial least squares( PLS) regression method was employed to construct the prediction model for chlorophyll content in rapeseed. The results of cross validation indicated that the coefficient of determination( Q2) of validation set was 0. 944 6,while the mean square error of cross validation( RMSECV) was 1. 36 mg/kg. The results indicated that this method could accurately determine chlorophyll content in rapeseed,which provided an important technological support to quality control of rapeseed oil.更多还原