Estimation of Breakdown Voltage of Epoxy/SiO2 Nanocomposites Insulators Using Machine Learning Algorithm

Several power utilities currently use non-ceramic insulators, like epoxy insulators. In order to improve the electrical properties, nano silicon dioxide (SiO
2
) was added to epoxy composite rod samples in this study. Testing for breakdown voltage was done. The breakdown voltage values for epoxy composites were estimated and predicted using machine learning algorithms (ML). The findings demonstrate that as sample length was increased and SiO
2
nanofiller was added, the breakdown of epoxy increased. The dispersion and distribution of the SiO
2
in the epoxy resin as well as the increased interfacial bonding were key factors in achieving the better characteristics of the nano SiO
2
filled composites. The epoxy sample with added SiO
2
nanoparticles of 20 mm length recorded the best breakdown voltage of 43.656 kV. Machine learning was able to calculate breakdown voltage values with very small percentage of error values, and was able to predict breakdown voltage values at sample lengths that were not fabricated.