Abstract: On 8 August 2017, the Jiuzhaigou M7.0 earthquake following the Wenchuan M8.0 earthquake and the Lunshan M7.0 earthquake was the third strong shock in the past 10 years in the eastern margin of the Qinghai-Tibetan Plateau. The Jiuzhaigou earthquake occurred in the intersection area among the Tazang, Minjiang and Huya faults in the eastern part of the Bayan Har block of the Qinghai-Tibet Plateau. In this paper, Time-frequency characteristics of the method is used based on EEMD decomposition extraction signal in the treatment of the Jiuzhaigou M7.0 earthquake records. Through EEMD decomposition, Hilbert transform, and spectral analysis on the acceleration records of stations near the epicenter, the time-frequency distribution characteristics of the signal are effectively obtained. The characteristics of the center frequency, Hilbert energy, and maximum amplitude are extracted quantitatively. And a comparative study between the Fourier transform and wavelet transform was made. The study shows that for the non-linear strong earthquake records, EEMD can be used to inhibit the existing problem of mode mixing of EMD decomposition. In low frequency, the amplitude of ground motion is underestimated. With the increase of frequency, FFT spectrum will magnify the amplitude.