In order to shorten the detection time of the electronic nose

a set of electronic nose system based on flow velocity modulation is designed

which uses variable intake velocity instead of fixed inlet flow

velocity and a kind of adaptive principal component analysis（AD—PCA） is put forward for classification.Beer classification is used to verify the system

and AD—PCA obtain an average correctness classification rate of 96.8 %. Compared with the fixed flow rate

this system can effectively shorten the detection time on the basis of ensuring accurate classification.