Abstract: Accurate extraction and analysis of harmonics is of great significance to power grid. To overcome the shortcomings of shorttime Fourier transform（STFT）and continuous wavelet transform（CWT）in harmonic analysis of complex power systems,Fourier-based synchrosqueezing transform（FSST）and synchrosqueezing wavelet transform（SWT）are proposed for harmonic analysis. These two methods respectively use STFT and CWT to analyze power harmonic signals,and synchronously extrude and sharpen the instantaneous frequency of harmonics to obtain more precise harmonic time-frequency curves;then use their reversibility to decompose power harmonics into a set of intrinsic mode type functions（IMTs）components,and complete the extraction of harmonic components. Through comparative study,it is found that FSST is more suitable for analyzing linear frequency modulation power harmonics,and SWT is more suitable for analyzing non-linear frequency modulation power harmonics. The practical application results show that compared with the Hilbert-Huang transform,the synchrosqueezing transform can effectively suppress the mode aliasing and noise interference,and extract the harmonic characteristic information effectively.