鸡肉中假单胞菌的近红外光谱快速识别

发布时间：2018-06-04 09:32

本文选题：鸡肉 + 近红外光谱　；参考：《农业机械学报》2017年08期

【摘要】：假单胞菌是鸡肉腐败最主要的致腐菌,为了快速识别鸡肉中的假单胞菌,首先从腐败鸡肉中分离并筛选出致腐菌,进一步利用聚合酶链反应技术对目标菌株进行生物学鉴别(分别为盖氏假单胞菌、嗜冷假单胞菌、莓实假单胞菌和荧光假单胞菌);配置鉴定的4种假单胞菌和等体积混合的4种假单胞菌菌液,采集菌液近红外透射光谱信息;然后运用标准正态变量变换对光谱进行预处理,利用联合区间偏最小二乘法筛选出特征波段;最后有比较地运用K最近邻法、最小二乘支持向量机和反向传播人工神经网络建立5种假单胞菌菌液的近红外光谱分类识别模型。其中反向传播人工神经网络模型预测效果最佳,其训练集和预测集的识别率分别为99.17%和95.00%。研究结果表明,近红外光谱结合反向传播人工神经网络可以快速识别鸡肉中的假单胞菌。
[Abstract]:Pseudomonas is the most important spoilage fungus in chicken. In order to quickly identify Pseudomonas in chicken, the spoilage fungi were isolated and screened. The target strains were further identified by polymerase chain reaction (Pseudomonas Gaysoni, Pseudomonas thermophilus). Pseudomonas raspberry and Pseudomonas fluorescens, four Pseudomonas species and four pseudomonas mixed with the same volume were configured to collect near-infrared transmission spectrum information, and then the spectrum was pretreated by standard normal variable transformation. The characteristic bands were screened by using the joint interval partial least square method, and the near infrared spectrum classification and recognition models of five pseudomonas liquid were established by using K-nearest neighbor method, least squares support vector machine and backpropagation artificial neural network. The back propagation artificial neural network model has the best prediction effect, and the recognition rates of training set and prediction set are 99.17% and 95.00% respectively. The results show that near infrared spectroscopy combined with back propagation artificial neural network can quickly identify Pseudomonas in chicken.
【作者单位】：江苏大学食品与生物工程学院;
【基金】：国家自然科学基金面上项目(31371770) 江苏省高校自然科学研究面上项目(16KJB550002) 江苏大学高级人才基金项目(15JDG169)
【分类号】：O657.33;TS251.7

【相似文献】