Conventional methods for on-line measurement of dielectric loss factor are difficult to meet the requirements of accuracy and anti-interference.A novel time-frequency analysis method for measuring dielectric loss factor is proposed in this paper.Hilbert-Huang transform（HHT） is adopted to process the voltage and current signals.Intrinsic mode functions（IMF） of voltage and leakage current signals are extracted by empirical mode decomposition（EMD）

and Hilbert transform is applied to calculation of the corresponding instantaneous frequency of fundamental components

then the tanδ can be calculated.The algorithm can accurately extract the fundamental components of voltage and leakage current without synchronous sampling.Simulation results show that there is little influence of sampled data length and frequency tracking error on HHT

and the proposed algorithm exhibits good application performance when the sampled data length is not exactly an integer fold of a power cycle

which demonstrates the algorithm can effectively improve the accuracy and stability of dielectric loss measurement.