【摘要】 采用国家标准GB∕T 6435—2016、GB∕T 6432—1994、GB∕T 6434—2006分别测定油菜籽的水分、蛋白质和纤维素,同时选择80份代表性的油菜籽样品建立油菜籽的水分、蛋白质和纤维素的近红外光谱仪检测数学模型。研究发现:用该模型测试35份油菜籽样品,所测水分数值与标准GB∕T 6435—2016所测数值进行比对,相关系数0.967,相对误差小于3.5%;用该数据模型测试50份油菜籽样品,所测蛋白质数值与标准GB∕T 6432—1994所测数值进行比对,相关系数0.993,相对误差小于3.0%;用该数据模型测试20份油菜籽样品,所测纤维素数值与标准GB∕T 6434—2006所测数值进行比对,相关系数0.954,相对误差小于3.0%。结果表明:三种数学模型的建立,可满足准确、快速、无污染检测油菜籽水分、蛋白质和纤维素的含量。更多还原

【Abstract】 Using the National Standard GB∕T 6435—2016,GB∕T 6432—1994 and GB∕T 6434—2006,themoisture,protein and cellulosecontent of rapeseed were respectively detected. Meanwhile, the mathematicalmodels of near-infrared spectrometer for detecting moisture,protein and cellulosecontent of rapeseed were builtusing 80 representative rapeseed samples. 35 rapeseed samples were tested with the model for moisture. Comparedwith the measurement results of moisture through the use of GB∕T 6435—2016,the correlation coefficient was0. 967 and the relative error was less than 3. 5%. In addition,50 rapeseed samples were tested with the model forprotein content. Compared with the measurement results of protein content by using GB∕T 6432—1994, thecorrelation coefficient was 0. 993 and the relative error was less than 3%. Besides,20 rapeseedsamples weretested with the model for cellulose content. Compared with the measurement results by using the GB∕T 6434—2006,the correlation coefficient was 0. 954 and the relative error was less than 3%. These results indicated thatthe establishment of the three kinds of mathematical models can meet the accurate, rapid and pollution-freedetection of moisture,protein and cellulose content in rapeseeds.更多还原