Breakdown strength estimation of air and a mixture of air+SF6 using artificial neural network approach

摘要

The breakdown strength of dry air and a mixture of dry air + SF6 containing 1 SF6 in non-uniform field was studied in a pressure range of 50 to 450 kPa. Addition of 1 of SF6 to dry air increased considerably the breakdown voltage in a non-uniform field. An Artificial Neural Network (ANN) gave a very good performance in the extrapolation of breakdown voltage values above the measuring range.