Abstract: To realize the rapid and accurate identification of sub/supersynchronous oscillation in power system, this paper proposes a sub/supersynchronous oscillation identification method based on synchrosqueezing generalized S transform (SSGST) and improved sparse time domain method (ISTD). The method uses the energy ratio function to detect the wide-area measurement information of the power system in real-time. When the signal energy is detected to be abrupt, the SSGST decomposes the detected oscillation signal to obtain the corresponding SSGST time-frequency coefficient matrix. Then, the improved ridge extraction method realizes the optimal trajectory search of each oscillation component in the time-frequency domain. Furthermore, the time-domain components of each oscillation component are reconstructed by combining the optimal trajectory time-frequency index, and the frequency and damping ratio coefficients of each oscillation component are calculated using the ISTD method. The self-synthesized signal, the simulation signal of the doubly-fed wind farm through the series compensation grid-connected system, and the measured data of an actual wind farm verify the accuracy and effectiveness of the proposed method.