【作者】 耿朝曦； 鲁超； 田磊； 王加华； 韩东海；

【Author】 Geng zhaoxi ,Lu chao,Tian lei,Wang jiahua,Han donghai (college of food science and nutritional engineering, China Agricultural University, Beijing 100083 )

【机构】 中国农业大学食品科学与营养工程学院；

【摘要】 本文考察了近红外光谱技术在酿酒葡萄白利糖度(Brix)检测方面的应用,所建模型经多元散射校正(MSC)、光谱分辨率的选择等预处理之后,其相关系数R都在0.95以上,预测标准值RMSEP在0.3以下,可以满足实际的需要;另外,本文还探讨了光谱的采集方式和分辨率对酿酒葡萄模型预测能力的影响。结果表明,对于酿酒葡萄这种特定的研究对象而言,它不受采集方式的限制;分辨率为16nm时模型的预测效果较好。更多还原

【Abstract】 This paper discussed the feasibility of near infrared spectrum technology for monitoring Brix of wine grapes. After pretreatment of Multiplicative Scatter Correction(MSC) and choosing different resolution, the correction coefficient of calibration(R)of the model was above 0.95 and it’s Root Mean Squared Error of Predication(RMSEP) was under 0.3. Otherwise, the paper also discussed the spectrum collection and the different resolution effect the prediction ability of the model. The reason indicated, for wine grapes, when it’s near infrared spectrum model is independent of the spectrum collection and the spectrum’s resolution is chose 16nm, the model is good. This paper was concluded that near infrared spectroscopy could be used as rapid alternative method for monitoring the maturity stage of wine grapes.更多还原