A portable electronic nose was applied in the detection and analysis of four groups of hundred-grass-oil with different production batches. Combined with the advantages of principal component analysis（ PCA） and linear discriminant analysis（ LDA）

PCA + LDA analysis method was adopted. By this method

the data collected was first analyzed by PCA to reduce the data dimension and optimize feature vectors

and then LDA was employed to classify the four sample groups. It worked out that the classification rate reached 87. 5 %

only a few samples of group p0705 and p0801 were classified incorrectly.