鸭肉中谷氨酸含量的可见-近红外光谱测定研究

为实现鸭肉中谷氨酸含量的快速测定,提出了利用可见-近红外光谱结合PCA、BP神经网络来建立鸭肉中谷氨酸含量测定的定量分析模型.采集试验首先采集光谱范围在350～1800nm的鸭肉可见-近红外反射光谱,并在430～1000、1001～1400和430～1400nm 3个光谱范围内分别用一阶导数(FD)、二阶导数(SD)、...

QUANTITATIVE ANALYSIS OF GLUTAMATE CONTENT IN DUCK MEAT USING VISIBLE-NEAR INFRARED SPECTROSCOPY

Quantitative analysis models of glutamate contents in duck meat were established using visible-near infrared spectroscopy combining with principal component analysis (PCA) and BP neural network to achieve the rapid prediction of glutamate contents in duck meat. Visible-near infrared spectroscopies of duck meat in the range of 350~1800nm were obtained, and first derivative(FD), second derivative(SD), multiplicative scatter correction(MSC) and standard orthogonal variable transformation (SNV) were used to complete the pretreatment of original spectroscopy in the spectral region of 430~1000, 1001~1400 and 430~1400nm, respectively. The scores of 8 principal components (PCs) and glutamate contents in duck meat were selected as the inputs and outputs of BP neural network to build the prediction model of glutamate content in duck meat, respectively. The results showed that the model of BP neural network established using the pretreatment spectroscopy by SNV in the spectral region of 430~1000nm was optimal. The correlation coefficient and the root mean squared error of prediction samples were 0.9564 and 0.058572 in the validation set respectively, and the model could meet the demands of rapid detection of duck meet glutamate content.