人工神经网络在鸡胸肉预冷清洗环节中沙门氏菌污染率的预测  

Application of Artificial Neural Network in Prediction of Salmonella Incidence in Chicken Breast Chilling Process

作　　者：肖兴宁 杨力 张建民[3] 廖明[3] 李延斌[4,5] 肖英平[1] 杨华[1] 汪雯[1] XIAO Xing-ning;YANG Li;ZHANG Jian-min;LIAO Ming;LI Yan-bin;XIAO Ying-ping;YANG Hua;WANG Wen(MOA Laboratory of Quality&Safety Risk Assessment for Agro-products(Hangzhou),State Key Laboratory for Managing Biotic and Chemical Threats to the Quality and Safety of Agro-products,Institute of Quality and Standard of Agricultural Products,Zhejiang Academy of Agricultural Sciences,Hangzhou 310021,China;College of Information Engineering,China Jiliang University,Hangzhou 310018,China;College of Veterinary Medicine,South China Agricultural University,Guangzhou 510642,China;College of Biosystems Engineering and Food Science,Zhejiang University,Hangzhou 310058,China;Department of Biological&Agricultural Engineering,University of Arkansas,Fayetteville,Arkansas 72701,USA)

机构地区：[1]浙江省农业科学院农产品质量标准研究所,农业农村部农产品质量安全风险评估实验室(杭州),农产品质量安全危害因子与风险防控国家重点实验室,浙江杭州310021 [2]中国计量大学信息工程学院,浙江杭州310018 [3]华南农业大学兽医学院,广东广州510642 [4]浙江大学生物系统工程与食品科学学院,浙江杭州310058 [5]阿肯色大学生物与农业工程系,阿肯色州费耶特维尔72701

出　　处：《食品工业科技》2020年第18期212-217,共6页Science and Technology of Food Industry

基　　金：沃尔玛基金会(SA1703164)。

摘　　要：为实现对鸡胸肉预冷清洗环节的沙门氏菌污染率的预测,采用响应面试验设计收集数据,建立以初始污染水平、初始污染率、次氯酸钠(NaClO)浓度为输入值,鸡胸肉预冷清洗环节的沙门氏菌污染率为输出值的广义回归神经网络模型(General Regression Neural Network model,GRNN),预测鸡胸肉预冷清洗环节的沙门氏菌污染率变化,并用训练集拟合,测试集评估模型的预测效果。结果显示,鸡胸肉预冷清洗环节的沙门氏菌污染率随初始污染水平、初始污染率的升高而显著增加,相反随NaClO浓度的升高而呈下降趋势(P<0.05)。练后的GRNN模型的r值和SEP值分别为0.93和10.8%,拟合良好。模型对新数据预测的误差较小(SEP=13%),表明GRNN模型可较准确的预测鸡胸肉预冷清洗环节的沙门氏菌污染率。本研究建立的模型可用于鸡胸肉预冷清洗环节沙门氏菌污染率的预测,为微生物定量风险评估提供重要信息。The prediction of Salmonella incidence in chicken breast chilling process was performed through the combination of response surface methodology(RSM)and general regression neural network(GRNN).RSM was utilized to collect the experimental data and the GRNN model was established to predict the changes of Salmonella incidence in chicken breast chilling process.In the GRNN model,the initial contamination level,pre-chill incidence and sodium hypochlorite(NaClO)concentration were considered as the input variable while post-chill incidence was regarded as the output variable.Furthermore,the training set was used for model fitting and the test set was used to evaluate the prediction ability of the model.The results showed that the post-chill incidence in chicken breast chilling process increased significantly with the initial contamination level and pre-chill incidence increased.On the contrary,post-chill incidence in broiler chilling process decreased significantly with the NaClO concentration increased(P<0.05).The GRNN model showed the best fit as indicated by the r(0.93)and SEP(10.8%)values.Besides,the model had a SEP value of 13%for new data,suggesting accurate prediction of the Salmonella incidence in chicken breast chilling process using GRNN.This study could use to predict Salmonella incidence in chicken breast chilling process at slaughter house and for quantitative microbial risk assessment as well.

关 键 词：鸡胸肉 预冷清洗 广义回归神经网络模型 沙门氏菌污染率 预测模型 

分 类 号：TS201.6[轻工技术与工程—食品科学]