摘要
利用自主研制的CNe-Nose I型电子鼻气体分析仪对多种品牌、不同新鲜程度的牛奶进行检测,并通过模式识别方法分析和识别数据.牛奶各组分经由金属氧化传感器阵列采集信号,其数据及响应曲线记录在PC端.运用主元分析和人工神经网络方法识别曲线特征,通过与新鲜高质量样本的标准数据进行比较,判断出牛奶变质程度.实验结果表明,电子鼻技术对牛奶品质的识别率较高,且具有便捷、安全等特点,是一种发展前景良好的实用技术.
A developed electronic nose system is described and used to measure different brands and qualities of milk. Pattern recognition algorithm serves for data analysis and recognition. In the experiment, milk components can be detected by Metal Oxide Sensor Array, and data graphs from sensors are then recorded in computer. Principal Component Analysis and Artificial Neural Network identify graph eigenvalues and distinguish different components. After, being compared to standard sample, milk's deteroprate grade can be estimated. Experimental results demonstrate good identification of this system in milk quality determination. With other advantages such as convenience and safety, electronic nose technology shows the potential in future aPPlication.
出处
《传感技术学报》
CAS
CSCD
北大核心
2007年第8期1727-1731,共5页
Chinese Journal of Sensors and Actuators
基金
浙江省国际合作重点项目资助(2004C24002)
关键词
电子鼻
牛奶检测
气体传感器阵列
主元分析
人工神经网络
electronic nose
milk detection
gas sensor array
principal component analysis
artificial neural network