基于Prony算法的电力电缆极化等效电路参数辨识

Parameter Identification of Equivalent Circuit for Power Cable Polarization Based on Prony Algorithm

为了解决当前XLPE电力电缆绝缘扩展Debye模型参数辨识中的问题,提出了一种基于Prony算法的参数辨识方法。首先利用去极化电流数据构建样本矩阵Re;然后根据样本矩阵Re的奇异值来确定扩展Debye模型弛豫支路数;最后用去极化电流数据构建新的样本矩阵R来进行扩展Debye模型的等效电路参数辨识。通过仿真和实测分析,证明该方法能有效地辨识扩展Debye模型的等效电路参数,且采用该方法辨识得到的最大时间常数可以作为表征热老化电缆绝缘老化程度的特征量。

In order to solve the problems in the parameter identification of extended Debye model for XLPE power cable insulation,a parameter identification method based on Prony algorithm was proposed.At first,the sample matrix Re was constructed using depolarization current data.Then,the relaxation branch number of extended Debye model was determined according to the singular value of the sample matrix Re.At last,a new sample matrix R was constructed using depolarization current data to conduct parameter identification for equivalent circuit of extended Debye model.According to the simulation and experiment analysis,the equivalent circuit parameters of extended Debye model can be identified effectively by the method,and the maximum time constant identified by this method can be used to characterize the ageing degree of thermal aged cable insualtion.