摘要
针对消费者在市场上购买到劣质休闲豆干产品后遇到维权困难等问题,建立了一种快速、准确的识别休闲豆干品牌真伪的方法。利用电子鼻对休闲豆干样品进行测量,获得198个样品的测量数据。随机选择140个作为训练集样品,剩余58个作为预测集样品。采用主成分分析(PCA)、偏最小二乘判别分析(PLS-DA)和误差反传人工神经网络(BP-ANN)等化学计量学方法处理样品的测量数据。结果显示:采用PCA方法,识别真伪品牌比较困难;采用PLS-DA方法,对真伪品牌进行建模,训练集的真品牌和伪品牌的识别率分别为100%和98.9%,预测集的真伪识别率分别为88.9%和100%;使用BP-ANN方法,真品牌和伪品牌的正确识别率均为100%。预测值与实际期望值高度吻合。采用电子鼻与BP-ANN方法相结合能够很好地识别休闲豆干真伪。该方法为休闲豆干的快速鉴别或厂家追溯提供了一定的参考和技术支持。
It is difficult for consumers to trace back to manufacturers and safeguard their rights after purchasing some leisure dried tofu products with quality problems in the market.In order to solve the problem,a rapid and accurate method was established to identify the true and false brand of leisure dried tofu.An electronic nose is sensitive to certain specific types of gas components and was used to measure the samples of leisure dried tofu.Data of 198 samples were obtained.140 samples were randomly selected from these samples as training set.The remaining 58 samples are used as prediction set.The chemometrics methods of back propagation artificial neural network (BP-ANN),partial least squares discriminant analysis (PLS-DA) and principal component analysis (PCA) were used to process the sample data.The results display that it was difficult to identify true and fake brands by PCA method.PLS-DA method was used to model the true and fake brands.The recognition rates of true and fake brands in training set were 100% and 98.9% respectively,and those of prediction set are 88.9% and 100% respectively.However,the BP-ANN method can gave a correct recognition rates of 100% for both true brand and fake brand.The predicted value was highly close to the actual expected value.Therefore,the combination of electronic nose and BP-ANN is an excellent way to identify the authenticity of leisure tofu.What is more,it provides some reference and support for quick identification of leisure dried tofu or factory tracing.
作者
杨莉
夏阿林
张榆
钱文菲
YANG Li;XIA Alin;ZHANG Yu;QIAN Wenfei(School of Food and Chemical Engineering,Shaoyang University,Shaoyang 422000,China)
出处
《食品科技》
CAS
北大核心
2020年第12期307-312,共6页
Food Science and Technology
基金
湖南省教育厅科学研究重点项目(16A236)
邵阳学院大学生研究性学习和创新性实验计划项目(2017CX75)
邵阳学院大学生创新创业训练计划项目(2019CX57)。
关键词
休闲豆干
电子鼻
化学计量学
判别分析
leisure dried tofu
electronic nose
chemometrics
discriminant analysis
