We designed an electronic nose system based on flow modulated,in the case of variable flow rates,in order to improve the accuracy of recognition and short the detection time, we maximize the gas response range of sensors to different components and concentrations by changing the intake flow rate.We used adaptive principal component analysis（AD-PCA） to classify the age of Chinese yellow wine in order to verify the system, and  compare the result of AD-PCA with support vector machine（SVM）and back-propagation neural network（BPNN）, The experimental resultshows that among the5 different ages of Chinese yellow wine,the average correct classification rate by AD-PCA is93.6%, SVM is92% and BPNN is 100%,it is proved that the system can quickly classifythewineage on the basis ofa higher accuracy rate, and can shorten the detection time compared with the fixed flow rate.