【摘要】 高粱作为酿造白酒的原料直接影响成品酒的品质,无损快速鉴别酿酒特需高粱的真实性是急需解决的重要问题。该研究针对高粱透光性差,近红外探头有效检测面积小,导致难以完全获取样品池中特需高粱和掺假高粱样品信息等问题,通过改良环形样品杯内径为略大于检测窗口直径的2倍,将高粱样品平铺于样品杯底部一层,解决了无法完全采集样品池中包括掺假高粱颗粒在内的所有样品信息等问题。结合化学计量学方法建立并优化了酿酒特需高粱真实性无损快速鉴别的近红外光谱模型,得到该模型的校正集、预测集和验证集的判别准确率分别为100%、100%和96.7%。更多还原

【Abstract】 Sorghum, as the raw material for Baijiu(Chinese liquor) production, directly affects the quality of the finished Baijiu, and it is an important problem to quickly identify the authenticity of the special sorghum for Baijiu production without damage. In this study, due to the poor light transmittance of sorghum and the small effective detection area of the near-infrared probe, it is difficult to completely obtain the information of specially needed sorghum and adulterated sorghum samples in the sample pool. By improving the inner diameter of the annular sample cup, which is slightly larger than the diameter of the detection window, the sorghum samples were tiled on the bottom layer of the sample cup, which solved the problem that all sample information including adulterated sorghum particles could not be collected completely in the sample pool. Combined with the method of stoichiometry, a nondestructive and rapid identification model of sorghum with special needs for Baijiu-brewing was established and optimized. The accuracy of calibration set, prediction set and verification set of the model were 100%, 100% and 96.7%, respectively.更多还原