The initial testing system made up of thirteen gas sensors

one temperature sensor and one humidity sensor is used to test three kinds of vinegars.Three-scale decomposition of the wavelet packet is employed to extract the feature values of measuring results

then principal component analysis（PCA） is applied to optimize sensor array.The optimized sensor array composed of five sensors is obtained finally.Using radial basis function neural network（RBFNN） to compare the identification effect between optimiz sensor array and original sensor array

the results of classification present that the optimized array can discriminate successfully three kinds of vinegars and improve the recognition performance compared with original array.