基于生物阻抗的即配羊肉货架期无损检测方法  

Bioimpedance-based Nondestructive Detection Method for Shelf-life of Ready-to-prepare Mutton

作　　者：李鑫星[1] 张子怡 梁步稳 黄晓燕 张国祥[1] 马瑞芹 LI Xinxing;ZHANG Ziyi;LIANG Buwen;HUANG Xiaoyan;ZHANG Guoxiang;MA Ruiqin(Beijing Laboratory of Food Quality and Safety,China Agricultural University,Beijing 100083,China;National Research Facility for Phenotypic and Analysis of Model Animals,China Agricultural University,Beijing 100083,China)

机构地区：[1]中国农业大学食品质量与安全北京实验室,北京100083 [2]中国农业大学模式动物重大设施建设办公室,北京100083

出　　处：《农业机械学报》2022年第7期379-386,共8页Transactions of the Chinese Society for Agricultural Machinery

基　　金：财政部和农业农村部:国家现代农业产业技术体系项目(CARS38);欧盟Switch Asia项目(DCI:ASIE/2012/307186)。

摘　　要：货架期是判断羊肉新鲜度的重要标准。为探讨生物阻抗技术在食品货架期检测方面的应用前景,提出了一种即配羊肉货架期无损检测方法。结合影响即配羊肉新鲜度变化的关键因素及生物阻抗的测量原理,针对电极数量、电极材料、电极排列方式等测试条件的不同,自主设计了电极作为生物阻抗测试前端。揭示了在0、4、8℃的3个贮藏温度下即配羊肉阻抗参数和TVB-N含量的变化规律及即配羊肉阻抗与TVB-N含量、货架期的相关性;以TVB-N含量为关键参考指标,建立基于BP神经网络的即配羊肉货架期预测模型和评价方法,并将其与支持向量机模型、决策树模型进行对比,BP神经网络模型的F1分数可达95.9%。基于BP神经网络模型设计即配羊肉货架期检测系统,可实现用户友好的数据可视化与即配羊肉货架期的即时检测。Shelf life is an important indicator for evaluating the freshness of mutton,which is directly related to its quality.To explore the application prospects of bioimpedance technology for shelf-life detection of food,a nondestructive and efficient shelf-life detection method was proposed for ready-to-prepare mutton.Combining the key factors affecting the change of freshness of ready-to-prepare mutton and the measurement principle of bioimpedance,the electrodes were designed independently for measuring bioimpedance according to the different testing conditions such as the number of electrodes,electrode materials,and electrode arrangement.The changes of impedance and TVB-N content of ready-to-prepare mutton at three storage temperatures of 0℃,4℃and 8℃and the correlation of impedance with TVB-N content and shelf life were revealed;a shelf-life prediction model and evaluation method of ready-to-prepare mutton based on BP neural network was established with TVB-N content as the key indicator,and it was compared with SVM(support vector machine)model and decision tree model.The F1-score of the BP neural network model was up to 95.9%.Based on the BP neural network model established above,a shelf-life detection system of ready-to-prepare mutton was developed by using Java language,which realized user-friendly data visualization and real-time detection of the shelf life of ready-to-prepare mutton.The research result can provide theoretical basis and software tool for the rapid and nondestructive detection of the shelf life of ready-to-prepare mutton,which can ensure the quality and safety of ready-to-prepare mutton and promote the sustainable and healthy development of the food industry.

关 键 词：即配羊肉 货架期检测 生物阻抗 BP神经网络 可视化系统 

分 类 号：TS207[轻工技术与工程—食品科学]