近红外光谱技术在农产品品质快速检测中的应用研究  

Research on the Application of Near-Infrared Spectroscopy Technology in the Rapid Detection of Agricultural Product Quality

作　　者：刘广志 LIU Guangzhi(Dong’e County Inspection and Testing Center,Dong’e 252200,China)

机构地区：[1]东阿县检验检测中心,山东东阿252200

出　　处：《食品安全导刊》2025年第11期84-87,共4页China Food Safety Magazine

摘　　要：随着农产品质量安全需求的提升,传统检测方法在时效性和成本上的局限性日益凸显。为验证近红外光谱技术在效率、成本和准确性方面的优势,本文以苹果、小麦和猪肉为对象,采用波长为780~2500 nm的近红外光谱仪采集数据,结合偏最小二乘法建立糖度、蛋白质及脂肪含量的定量预测模型。结果显示,苹果糖度模型表现出最佳精度,相关系数达0.98,校正均方根误差为0.07%,验证集绝对偏差0.07°Bx,可满足快速无损检测需求;小麦蛋白质模型校正相关系数为0.90,校正均方根误差为0.12%,性能稳定;猪肉脂肪模型受肌内脂肪分布不均影响,精度略低。以上结果说明,近红外光谱技术在水果和谷物成分检测中具有高适用性,但在肉类检测中需优化模型。相较于传统化学方法,近红外光谱技术可提升检测效率,降低成本,且可实现无损检测。With the improvement of the demand for the quality and safety of agricultural products,the limitations of traditional detection methods in terms of timeliness and cost have become increasingly prominent.In order to verify the advantages of near-infrared spectroscopy in terms of efficiency,cost and accuracy,this paper used a near-infrared spectrometer with a wavelength of 780~2500 nm to collect data from apples,wheat and pork,and combined with the partial least squares method to establish a quantitative prediction model of sugar content,protein and fat content.The results show that the apple sugar content model shows the best accuracy,with a correction coefficient of 0.98,a root mean square error of 0.07%,and an absolute deviation of 0.07°Bx in the validation set,which can meet the needs of rapid nondestructive testing.The adjusted correlation coefficient of the wheat protein model was 0.90,and the root mean square error was 0.12%,indicating stable performance.The pork fat model was affected by the uneven distribution of intramuscular fat,and the accuracy was slightly lower.The above results show that near-infrared spectroscopy has high applicability in the detection of fruit and grain components,but the model needs to be optimized in meat detection.Compared with traditional chemical methods,near-infrared spectroscopy can improve detection efficiency,reduce costs,and achieve non-destructive testing.

关 键 词：近红外光谱 农产品品质 化学计量学 快速检测 

分 类 号：O657.3[理学—分析化学]