论文部分内容阅读
以完整甘蓝型油菜籽为样品,研究了不同光谱预处理和回归统计方法在用近红外反射光谱（NIRS）分析油菜籽中芥酸和硫甙的含量时,对建立回归方程的影响。结果表明,光谱预处理对校正结果影响较大,不同光谱数学处理以2阶导数处理较好,在各阶导数处理中存在着交叉现象,对于芥酸和硫甙都可找到一些较好的处理组合。对于芥酸,除趋势变换法（De-trending）散射处理外,其余各处理均有明显效果。但散射处理对硫甙分析效果不如芥酸明显。回归统计方法对建立芥酸和硫甙的回归方程影响最为有效,两者表现一致,其中以改良偏最小二乘法（MPLS）效果最好。对于芥酸和硫甙,采用“标准正态变量转换（SNV）+趋势变换法（De-trending）/2,4,4,1/MPLS”的组合建立回归方程效果较好,检验平均偏差（Bias）分别小至-0.405和0.315,而检验决定系数（RSQ）分别高至0.982和0.972。本研究采用整粒小样品（3g）来分析,效果较好,可直接用于育种早世代选择。
Taking intact rapeseed (RAPD) as sample, the effects of different pretreatment and regression methods on the establishment of regression equation were studied when the content of erucic acid and glucosinolate in rapeseed was analyzed by near infrared reflectance spectroscopy (NIRS). The results show that the spectral preprocessing has a great influence on the calibration results. Different spectral mathods are better than the second derivative. There is a crossover phenomenon in the derivative of each order, and some better treatments can be found for erucic acid and glucosinolate combination. For erucic acid, except for the De-trending scattering treatment, the rest of the treatments have a significant effect. However, the scattering treatment of glucosinolates less effective than erucic acid. The regression statistical method has the most effective effect on establishing the regression equation of erucic acid and glucosinolate, and the two are consistent. The improved partial least squares (MPLS) is the best. For erucic acid and glucosinolate, using the combination of “SNV + De-trending / 2,4,4,1 / MPLS”, the regression equation is better and the mean deviation ( Bias were as small as -0.405 and 0.315, respectively, while the test determination coefficient (RSQ) was as high as 0.982 and 0.972, respectively. In this study, the whole small sample (3g) to analyze, the effect is better, can be used directly for breeding early generations to choose.