摘要
This paper presents new approach to using Partial Discharge (PD) data for the prediction of the Remaining Useful Life (RUL) of dielectric materials undergoing breakdown. The method presented uses a thermodynamic macro-model in conjunction with an artificial neural network to associate the features in the PD data detected during breakdown to the electrical tree characteristics. The method is presented using electrical tree simulation data from a new dielectric breakdown simulation model. The simulation model is confirmed using experimental data.
- 出版日期2015-2