宁夏玉米营养成分近红外分析模型建立与应用

研究采集宁夏区内各市县(固原、海原、同心、中卫、中宁、永宁、平罗)抽取玉米样品120批,用近红外分析仪扫描定标样品集获得玉米近红外光谱图,利用偏最小二乘法建立模型,并分别经过无预处理、均值中心化、标准正态变量转换、一阶导数、标准正态变量转换结合去趋势校正(SNV+D)预处理光谱,获得水分、粗蛋白、粗脂肪、粗纤维预测模型,分别是经过MC、SNV+D、SNV+D、SNV+D光谱预处理后效果最好。预测决定系数R² _p分别为0.951、0.976、0.728、0.897,相对分析误差RPD除粗脂肪为2.62外,其他均大于3的近红外预测模型,通过预测模型验证集验证后,并对模型预测值与实测值进行U检验,结果为差异不显著(P>0.05)。该近红外预测模型对玉米中水分、粗蛋白、粗纤维具有较佳预测效果,粗脂肪的预测精度有待进一步提高。

Near Infrared Analysis of Maize Nutrient Composition in Ningxia Establishment and Application of the Model

This study collected in various cities and counties in ningxia area(Guyuan、Haiyuan、Tongxin、Zhongwei、Zhongning、Yongning、Pingluo)evalution,haiyuan,concentric,centre-half,ning,yongning,PingLuo)in 120,corn samples drawn using near infrared analyzer collecting spectral scanning calibration set,using Partial Least squarescalibration model is set up,and after the Mean without pretreatment,centralized,Standard Normal variables transformation(Standard Normal Variable,FD and standard normal variable conversion combined with detrend correction(SNV+D)pretreatment spectrum obtained the prediction model of water,crude protein,crude fat and crude fiber,which obtained the best effect after MC,SNV+D,SNV+D,and SNV+D pretreatment spectrum respectively.The predicted determination coefficient R2 p was 0.951,0.976,0.728 and 0.897 respectively,and the Relative percent deviation of RPD was larger than 3 except that crude fat was 2.62.After verification by the prediction model validation set,the U test was carried out on the predicted value and the measured value of the model,and the results showed no significant difference(P>0.05).The near infrared prediction model has better prediction effect on water,crude protein and crude fiber in corn,and the prediction accuracy of crude fat needs to be further improved.