传感技术学报2011，Vol.24Issue(1)：10-13,4.DOI:10.3969/j.issn.1004-1699.2011.01.003
不同产地和采收期的中药材电子鼻鉴别研究
Discrimination of Chinese Herbal Medicines from Different Growing Areas and Harvesting Dates Based on E-Nose*
摘要
Abstract
It is difficult to distinguish the same kind of Chinese herbal medicines (CHMs)with different growing areas and harvesting dates. An electronic nose(E-nose) PEN3 was used to discriminate the CHMs with different growing areas and harvesting dates under natural state. Firstly, odor information of samples was obtained by E-nose, and feature set was constituted by all aspects of the representative characteristics of odor information. Then, negative feedback was applied to extract features, and the most effective feature subset was determined. PCA+LDA was used to classify samples, finally, unknown samples were identified by using Euclidean distance, Mahalanobis distance. Identification of unknown samples was up to 100％ accuracy, showing that it is feasible to evaluate the quality of the CHMs used e-nose. The paper providesa new method for quality control of CHMs.关键词
电子鼻/中药材/特征提取与选择/PCA+LDA算法/质量控制Key words
electronic nose/ Chinese herbal medicines/ feature extraction and selection/ PCA + LDA algorithm/quality control分类
信息技术与安全科学引用本文复制引用
伍世元,骆德汉,邓炳荣,赵庆辉..不同产地和采收期的中药材电子鼻鉴别研究[J].传感技术学报,2011,24(1):10-13,4.基金项目
2009年国家自然科学基金资助"基于仿生嗅觉的辛味中药材气味指纹图谱研究"(60971105) （60971105）
