In view of the problem that the long measurement of recovery voltage polarization spectrum and frequency domain dielectric spectrum affect identification efficiency of the extended Debye equivalent circuit parameters of transformer oil-paper insulation system, this paper proposes a novel identification method. At first, this paper proposes characteristics of hybrid time-frequency dielectric spectrum, which consist of the recovery voltage polarization spectrum characteristics under short charging time and small amount of frequency dielectric spectrum characteristics. Then, based on these proposed characteristics, using multi-objective particle swarm optimization (PSO) to identify equivalent circuit parameters. At last, compare the test time of the proposed method with the on-site testing. The results show that: equivalent circuit parameters of the extended Debye constructed by using characteristics of the hybrid time-frequency dielectric spectrum can effectively restore the complete recovery voltage polarization spectrum and frequency domain spectrum which field measured. The coincidence between the calculated and measured values of the two kinds of dielectric spectrum is above 95%. Moreover, the new method can reduce the test time by 61% compared to the recovery voltage measurement alone and by 46% compared to the frequency domain spectrum alone. The new method not only ensures the accuracy of parameter identification, but also greatly reduces the test time and improves the test efficiency. It is of great significance to the application and popularization of dielectric spectrum measurement technology in transformer on-site insulation detection.