可见-近红外光谱用于鲜食葡萄感官偏好的检测

发布时间：2019-03-08 09:57

【摘要】：水果内部品质影响着消费者的感官偏好,直接决定消费者的购买倾向。为此,可见-近红外光谱技术检测巨峰葡萄内部品质以及感官偏好等级。试验比较了漫反射、透射两种光谱采集方式对葡萄内部品质的检测准确性,结果表明透射光谱更能表征巨峰葡萄的内部品质信息,偏最小二乘(PLS)模型对鲜食葡萄可溶性固形物、总酸度的预测均方根误差分别为0.598%brix,0.048g·L~(-1)。将鲜食葡萄的透射光谱主成分信息与消费者的感官偏好等级之间建立非线性分类模型,得到主成分法结合极限学习机(PCA-ELM)模型分类准确率最好,为78.7%。结果认为反映水果内部信息的光谱可用于消费者感官偏好的初步分类,但其间关系还有待进一步研究。
[Abstract]:The internal quality of fruit affects the consumer's sensory preference and directly determines the consumer's purchase tendency. Therefore, visible-near infrared spectroscopy (NIR) was used to detect the internal quality and sensory preference grade of Jufeng grape. The accuracy of detecting the internal quality of grape by diffuse reflection and transmission spectra was compared. The results showed that the transmission spectrum could better characterize the internal quality information of Jufeng grape. The root mean square error (RMS) of partial least squares (PLS) model was 0.598% brix and 0.048g 路L ~ (- 1) for the prediction of total acidity of fresh grape soluble solids, and the root mean square error was 0.598% Brix and 0.048g 路L ~ (- 1) respectively. A nonlinear classification model was established between the transmission spectrum principal component information of fresh grape and the consumer's sensory preference level. The classification accuracy of the principal component method combined with the limit learning machine (PCA-ELM) model was 78.7%. The results suggest that the spectra reflecting the internal information of fruits can be used for the preliminary classification of consumers' sensory preferences, but the relationship between them needs to be further studied.
【作者单位】：江苏大学食品与生物工程学院;
【基金】：国家科技支撑项目(2012BAD29B04-4) 江苏省普通高校研究生科研创新计划(1291360015)资助
【分类号】：TS255.7;O657.33

【相似文献】