Predicting TJ Electric Discharge in Insulators by Developing, Comparing then Validing of Regression Models

EasyChair Preprint no. 13149

Abstract

The discharge and breakdown characteristics caused by triple junctions (TJ) in high-voltage (HV) electrical insulation are becoming increasingly important problems in recent industrial applications. To better organize the tests supporting our study and highlight the parameters and their interactions, we have developed a mathematical regression model of the monovariable polynomial discharge current as a function of the inter-electrode distance (d) and the applied voltage (V) independently and a two variable quadratique model as a function of d and V. The latter can simulate the maximum discharge current (I₀) with an adjusted determination coefficient close to 0.99. Therefore, the dual variable quadratic model was tested and validated.

Keyphrases: electric field, Electrical Discharge Current, model, Regression, triple junction

@Booklet{EasyChair:13149,
  author = {Nabila Saim},
  title = {Predicting TJ Electric Discharge in Insulators by Developing, Comparing then Validing of Regression Models},
  howpublished = {EasyChair Preprint no. 13149},

  year = {EasyChair, 2024}}