基于近红外光谱仪分析中式爆炒猪肉的水分含量

摘 要：应用傅里叶变换近红外光谱技术建立中式传统爆炒猪肉片水分含量预测模型，达到快速无损检测的目的。通过直接干燥法测定100 组爆炒猪肉片样品的水分含量，并扫描得到其近红外光谱图。采用偏最小二乘法并通过二阶微分结合卷积平滑算法对光谱进行预处理，通过拐点法、马氏距离法、杠杆值、学生残差法与内部交互验证均方根法进一步剔除异常样本，优化光谱模型。结果表明：所构建的中式爆炒肉片水分含量近红外光谱预测模型的校正均方差值为0.089 1，相关系数为0.972 1；且将预测值与真实值进行比较发现，预测结果正确率大于

98.7%（P<0.05），表明本研究建立的用于检测中式爆炒肉片水分含量的近红外光谱预测模型效果良好，能够快速检测并准确预测中式爆炒肉片的水分含量，具有一定的应用价值。

关键词：近红外光谱；中式爆炒猪肉；水分含量；无损检测

Detection of Water Content of Chinese Stir-Fired Pork by Near Infrared Spectroscopy

ZHAO Juyang， SHI Changbo*， FANG Weijia

（College of Tourism and Cuisine， Harbin University of Commerce， Harbin 150076， China）

Abstract： In the study， a nondestructive and rapid method for detecting the water content of Chinese stir-fired pork was established by near infrared reflectance spectroscopy. The moisture contents of 100 groups of stir-fired pork samples were determined by direct drying method and near infrared spectra of the samples were acquired. The original spectra were pretreated by second derivative and Savitzky-Golay smoothing in order to perform a partial least squares regression. Then inflection point， Mahalanobis distance， root mean square error of cross-validation and studentized residual were used to eliminate the outliers in order to establish a calibration model. The results showed that the proposed predictive model had a high correlation coefficient （r = 0.972 1） with a root mean squared error of calibration of 0.089 1. Furthermore， the prediction accuracy was greater than 98.7% （P < 0.05） when compared with the true value indicating the model established in this study has a good predictive performance. The results showed that the model allows accurate prediction of water content and has the potential for use in Chinese stir-fired pork processing and food industry.

Keywords： near infrared reflectance spectroscopy； Chinese stir-fired pork； moisture content； nondestructive detection

DOI：10.7506/rlyj1001-8123-201807008

中图分类号：TS207.3 文献标志码：A文章编号：1001-8123（2018）07-0042-07

引文格式：

赵钜阳， 石长波， 方伟佳. 基于近红外光谱仪分析中式爆炒猪肉的水分含量[J]. 肉类研究， 2018， 32（7）： 42-48. DOI：10.7506/rlyj1001-8123-201807008. http：//www.rlyj.pub

ZHAO Juyang， SHI Changbo， FANG Weijia. Detection of water content of chinese stir-fired pork by near infrared spectroscopy[J]. Meat Research， 2018， 32（7）： 42-48. DOI：10.7506/rlyj1001-8123-201807008. http：//www.rlyj.pub

爆炒是中国传统典型烹饪工艺之一，爆炒主要通过热对流、热传导和热辐射的方式进行[1]，由于烹饪温度高、时间短、过程激烈，因而在爆炒过程中，食品顆粒表面温度会迅速升高，但食品颗粒中心点达到成熟时，水分还未来得及向食品表面流动和扩散，整个爆炒过程已经完成，因而爆炒的菜肴脆嫩爽口[2]。然而，传统的爆炒方法往往由于厨师的经验性人为操作而造成炒制时间过度，进而影响肉品的多汁性、嫩度和口感[3]，因此急需对中式烹饪肉类菜肴的品质进行测定。