Electrical transformers are considered one of the very
important elements on the operation of power systems. The
operation of electrical transformers is dependent on the
performance of transformer oils. Hence, it is very necessary for
continuous and online checking the transformer oil insulation
level. Transformer oil wet process generates oil discharge space
charge in the long run and bubbles are formed. When the
electric field reaches a certain limit, partial discharge (PD)
appears and lead to deterioration of oil insulation level. In this
paper, Multi-wall carbon nanotube (MWNT) sensor is used for
detecting PD of oils). The performance of (MWNTs) films
sensor for continuous and on line monitoring the oil insulation
level is studied. A linear, fuzzy and neural regression models
are used to explain and predict the performance of the sensor.