【摘要】 论述了采用近红外光谱仪DA7200检测冻玉米中主要成分含量的方法。讨论了采用偏最小二乘法(PLS)建立校正模型过程中样品预处理及利用常规吸收峰优选波长的方法。经验证:水分、淀粉测量值同浓度参考值具有良好相关性(相关系数大于0.9),测量重复性变异系数(CV)优于2%。结果表明,近红外光谱法可以满足冻玉米中主要成分的实际测量要求,为玉米深加工企业提供了冬季原料玉米"按质收购"的参考方法。更多还原

【Abstract】 The method for assaying the contents of the main constituents of the frozen maize with DA7200 near-infrared spectroscope is presented. The sample pre-treatment during setting up the correcting model by partial least-squares (PLS) and the way of optimizing wavelength according to normal absorption peak are discussed. It is verified that there is a good correlation between the measured moisture and starch values and the reference value of concentration (the correlation coefficient is higher than 0.9), with a coefficient of variation (CV) of repeatability below 2%. Therefore, NIR spectroscopy can meet the practical measuring accuracy as required for the principal constituents of the frozen maize, thus providing the maize further-processing enterprises a reference for purchasing their raw material according to its quality in winter.更多还原