【摘要】 研究了电荷耦合器件(CCD)近红外光谱仪(NIRS)在短波区域(700￣1100nm),测定牛奶中蛋白质、脂肪、乳糖等三种常规成分的含量。采用散射校正(MSC)和偏最小二乘回归(PLS)相结合的方法建立了定量模型。将短波NIRS对未知样品的预测结果与FOSSOMATIC4000在的测量结果进行比较,对预测的准确性进行了考察。更多还原

【Abstract】 A charger-coupled-device-near-infrared spectrometer (CCD-NIR) was used to determine the content of protein, fat and lactose et al in milk in short wave near infrared spectral region (700￣1100nm). The qualitative calibration was developed by multiple scatter correction combine with partial least square regression. The result predicted by short wave NIRS to unknown sample was compared with that by FOSSOMATIC 4000 in long wave spectral region, the accuracy was studied.更多还原