BREAKDOWN OF COMPOSITE LIQUID VEGETABLE OIL AS DIELECTRIC A SIMULATION APPROACH

Abstract

This report deals with the development of the techniques for finding the breakdown voltages of the different mixture ratios of composite vegetable oil as a dielectric used in transformer without actually performing the experiments in the laboratory and finding the relationship between inputs and outputs of the system using artificial neural network model and to find the errors. Almost every aspect involved in the system modelling for conducting the research work has been explained. The main objective of the project is to study and carry out the assessment how the ANN model is trained and best model is created for finding the predicted output and also to find out the errors and in the next part of the dissertation we have extended the simulation work to find out the breakdown voltage of different mix ratios of composite oil by varying the different input parameters (one by one) respectively.