摘要
尝试采用电子鼻检测手段,对两种不同甲醛含量的香菇进行快速检测与识别。以两种不同甲醛含量的香菇为研究对象,利用主成分分析(PCA)、BP神经网络对电子鼻采集的数据进行分析识别。结果表明,主成分分析(PCA)能够很好的进行识别区分,用BP神经网络进行分析,训练样本回判率达到100%,测试样本识别率达到90%以上。电子鼻可以识别两种不同甲醛含量的香菇。
A rapid detection was made to determine two Lentinus edodes with different formaldehyde concentrations using an electronic nose （ Enose）. Two Lentinus edodes with different formaldehyde concentrations were studied in the experiment. The data obtained by electronic nose （ Enose） were analyzed by the method of principle components analysis （PCA） and article neural network （ANN）. Experimental result showed that the two shiitake mushrooms could be distinguished well by the method of principle components analysis （PCA）. Then the data were processed using article neural network （ANN）, in which the 100% correct classification from training samples and more than 90% identification ratio from testing samples were achieved ; E-nose could rapidly detect two Lentinus edodes with different formaldehyde concentrations.
出处
《食品与机械》
CSCD
北大核心
2009年第3期101-102,共2页
Food and Machinery
基金
浙江省教育厅高校优秀青年教师资助项目
浙江林学院校基金项目(编号:2005FR039)
关键词
香菇
甲醛
电子鼻技术
Lentinns edodes
Formaldehyde
Electronic nose technology
作者简介
庞林江（1977-），男，浙江林学院农业与食品科学学院讲师。E—mail：ljpang@zjfc．edu．cn
通讯作者：王允祥
