基于贝叶斯网络的猪肉合格率的模型构建  

Modeling the qualifying criteria of pork based on Bayesian network

作　　者：张方怡[1] 董庆利[1] 黄宋琳[1] 黄勋娟[1] 王文迪[1] 

机构地区：[1]上海理工大学医疗器械与食品学院,上海200093

出　　处：《食品工业科技》2012年第10期52-54,93,共4页Science and Technology of Food Industry

基　　金：国家自然科学基金(30800864);上海理工大学微创微创励志创新基金

摘　　要：我国是猪肉消费大国,其卫生情况对人民关系重大。近年来肉制品的微生物污染现象不容乐观,所以建立一个行之有效的猪肉追溯系统有利于更好地控制猪肉质量。本文根据现有的猪肉卫生标准,选出猪肉的四种优势致腐菌:大肠杆菌、单增李斯特菌、沙门氏菌以及金黄色葡萄球菌,以贝叶斯网络为理论依据,利用Hugin Lite软件拟合出合格率模型。结果表明,基于贝叶斯网络可通过改变菌体数据快速求得猪肉的合格率,大大缩减检验时间,为猪肉卫生合格率的质检工作提供理论依据。A significant proportion of China＇s meat consumption consists of pork.Therefore the issue of pork food safety is a matter of public concern.The establishment of an effective meat traceability system is necessary to monitor food safety.Using the Bayesian network theory to build the model by the Hugin Lite software system,we had identified the qualifying criteria in relate to four specific spoilage organisms（SSOs）,including E.Coli,Listeria monocytogenes,Salmonella and Staphylococcus aurous.By using this model,the qualifying criteria could be examined efficiently even when inputting variable bacteria levels.

关 键 词：贝叶斯网络 肉 追溯模型 

分 类 号：TS251.1[轻工技术与工程—农产品加工及贮藏工程]