Abstract: After asynchronous interconnection, the mixed oscillation of low frequency and ultra-low frequency will appear in the system with high permeability of hydropower units, which will do great harm to the safe and stable operation of power grid. To solve the problem of inaccurate identification of characteristic parameters of mixed oscillation signals, a second-order synchronous extracting wavelet transform (SSEWT) method is proposed, which is combined with empirical envelope (EE) demodulation technology to identify real-time modal parameters of oscillation signals. The SSEWT extends the existing synchronous extraction transform to wavelet transform, and introduces the second-order instantaneous frequency estimator to decompose the signal. Then, combined with EE demodulation technology, the parameters of each group of oscillation components are identified. Finally, the feasibility and accuracy of the proposed method are proved by self-synthesis signal simulation, 10-machine 39-bus system simulation, and measured signal simulation. Compared with empirical mode decomposition (EMD) algorithm and Fourier synchrosqueezed transform transform(FSST) algorithm, the simulation results show that the proposed method can be adopted to effectively suppress noise and endpoint effect, and accurately identify the characteristic parameters of mixed oscillation signals.