CAJ | 학술논문

试验采用不同波段的近红外光谱对原料奶的主要成分进行品质检测。使用2种近红外光谱仪采集原料奶的透反射和漫反射光谱，建立牛奶中蛋白含量、脂肪含量和乳糖含量的定量分析模型。结果表明，蛋白含量、脂肪含量、乳糖糖含量的相关系数（r）分别达到o．9；5¨、o．9218、o．8288，预测误差均方根（P,,MSEP）分别为1．g144、2．01们、2．804，测量值与浓度参考值之间有着良好的相关性。．结果表明，基于近红外光谱的原料奶主要成分品质快速检测准确度高，具有很高的实用价值。
시험채용불동파단적근홍외광보대원료내적주요성분진행품질검측。사용2충근홍외광보의채집원료내적투반사화만반사광보，건립우내중단백함량、지방함량화유당함량적정량분석모형。결과표명，단백함량、지방함량、유당당함량적상관계수（r）분별체도o．9；5¨、o．9218、o．8288，예측오차균방근（P,,MSEP）분별위1．g144、2．01문、2．804，측량치여농도삼고치지간유착량호적상관성。．결과표명，기우근홍외광보적원료내주요성분품질쾌속검측준학도고，구유흔고적실용개치。
Different wavelengths of near-infrared spectroscopy were used to test the quality of main components of raw milk. The quantitative an alysis model of protein content, fat content and lactose content was established by 2 kinds of near-infrared spectroscopy based on reflection and diffuse reflection spectra. The results gave the correlation coefficients of 0.9311, 0.9218, 0.8288 for protein, fat and lactose, respectively, with the root mean square error of prediction（RMSEP） of 1.9144, 2.0143, and 2.804. The measurements and concentration had a good correlation.The results demonstrated that near-infrared spectrometry was a valuable, rapid and nondestruction tool for the quantitative analysis of the principle components in row milk.