The identification system of impact location is constructed by a fiber Bragg grating (FBG) sensor network, and the wavelet transform, the frequency spectral analysis and support vector classifiers algorithm are used to identify the impact location. According to the impact test on the division area, the relationship between the impact area and signal characteristics is explored. On the basis of using wavelet transform to remove the low velocity impact signal baseline interference, the identification method of low velocity impact area location based on signal amplitude frequency characteristics is proposed. Support vector classifiers machine whose input is signal amplitude frequency characteristics and output is impact area to realize the identification of low velocity impact area location is built. The result shows that for 36 impact area on the 500 mm×500 mm×2 mm carbon fiber reinforced plastics plate under test, 33 impact area is accurate identification, regional location accuracy is above 90%, regional location accuracy is 40 mm×40 mm and each area identification time is less than 1011 ms. The research results provide a scientific and reliable method for the identification of the carbon fiber composite material plate location.