An Improved Broadband Common-Mode Electrical Machine Model for Online Condition Monitoring of Stator Insulation Degradation

Dayong Zheng,Pinjia Zhang
DOI: https://doi.org/10.1109/ecce.2019.8912863
2019-01-01
Abstract:Online condition monitoring for stator insulation is crucial to the reliability of AC electrical machines. Common-mode (CM) monitoring technique becomes popular in recent years. However, the CM models used in the existing monitoring methods are usually lumped-parameter models. The parameters in these models are usually obtained by curve-fitting. Thus these parameters lack specific physical meanings and these models are difficult to be used to study different insulation ageing problems. In addition, the parameters in these models are usually considered constant rather than frequency-dependent. This problem largely affects the accuracy of the model in broadband spectrum. In this paper, an improved detailed CM model for insulation monitoring is proposed. The electrical parameters in this model are obtained by calculation based on the geometrical and dielectric parameters of the winding insulation rather than curve fitting. The frequency-dependent parameters are calculated by finite element analysis (FEA). The proposed model is validated by experiments. The model can be used for analysis of CM online stator insulation monitoring technique in broadband spectrum.
What problem does this paper attempt to address?