The construction of the equivalent circuit model of power cable insulation is critical to study the evaluation of the aging status of power cable insulation

and the identification of the extended Debye equivalent circuit parameters is a challenging issue in this part. In view of the problem that the result of the traditional equivalent circuit parameter identification method is greatly affected by the time constant of each branch

this paper presents a new method based on singular entropy and total least squares-estimation of signal parameters via rotational invariance techniques（TLS-ESPRIT） algorithm to identify the equivalent circuit parameters of extended Debye model. First

the curve of the singularity entropy increment is created by the depolarization current

and the number of polarization branches of the extended Debye equivalent circuit is obtained by the curve of singularity entropy increment. Then the time constant and relaxation coefficient of each branch are derived by the TLS-ESPRIT algorithm. Finally

the resistance and capacitance of each branch are obtained. The simulation and experimental results show that the proposed technique can effectively obtain the number of polarization branches of the extended Debye equivalent circuit

the time constant and relaxation coefficient of each branch can be calculated uniquely

and the result is not affected by the time constant of each branch.