论文部分内容阅读
运用近红外光谱技术,开展带鱼贮藏过程中新鲜度研究,建立一种基于近红外技术的带鱼新鲜度快速检测方法。带鱼在4、0℃贮藏下原始光谱经一阶导数+矢量归一化处理后,进行聚类分析,并结合感官评定和挥发性盐基氮(TVBN)进行综合评价;通过对数据预处理,模型优化,用偏最小二乘法(PLS)建立TVBN含量和近红外光谱模型并对未知样品进行含量预测。结果表明:聚类分析结果与感官评定、TVBN含量一致;建立TVBN含量和近红外光谱模型的决定系数R2和预测均方根误差RMSECV分别为98.19%和1.95%,对未知样品预测的相对误差均小于5%。此方法快速简便、准确可靠,可用于带鱼新鲜度的判定。
Near infrared spectroscopy was used to study the freshness of the octopus during storage and to establish a rapid detection method for the freshness of octopus based on near infrared spectroscopy. The original spectra of the octopus were stored at 4 ℃ and the original spectra were normalized by the first derivative and vector normalization. The clustering analysis was carried out and the sensory evaluation and the TVBN were combined to evaluate the original spectra. After preprocessing, The model was optimized. TVBN content and near-infrared spectroscopy model were established by partial least squares (PLS) and the content of unknown samples was predicted. The results showed that the clustering results were consistent with the sensory evaluation and TVBN content. The determination coefficients R2 and RMSECV of establishing TVBN content and near-infrared spectral model were 98.19% and 1.95%, respectively. The relative error of predicting the unknown samples were Less than 5%. This method is fast and easy, accurate and reliable, which can be used to determine the freshness of octopus.