摘要
用四单元金属氧化物传感器阵列构成电子鼻系统,采用运算复杂度较小的特征提取与模式识别方法实现了三种卷烟的鉴别,并通过一系列比较实验研究了特征提取方法、气室、卷烟样品温度以及载气流量等对识别结果的影响。
Three different brands of cigarettes are identified using an electronic nose with an array of four SnO_2 sensors and algorithms of less complexity for feature extraction and pattern recognition. The effects of feature extraction methods, array chamber dimension and structure, cigarettes sample temperature and carrier gas flow on the identification results are discussed.
出处
《传感器技术》
CSCD
北大核心
2004年第6期62-65,共4页
Journal of Transducer Technology
关键词
电子鼻
卷烟鉴别
特征提取
模式识别
electronic nose
identification of cigarettes
feature extraction
pattern recognition
