Analysis of Distribution Transformer Physiological and Electrical Fault Detection - A Smart Grid Application

Abstract

Power grids transport electricity from the point of generation to the market. Power conversion from HV to LV and vice versa occurs in grids, also known as substations. These substations or power grids can be accessible or situated in remote locations. Transformers are used to convert power; they are an essential part of transmission and distribution networks. The method of grid monitoring and maintenance is essentially a very monotonous one. Monitoring the health of the transformers to maintain an uninterrupted power supply to the customers is difficult in such circumstances. Overvoltage, load currents, oil temperature, transformer oil level, and other parameters are monitored. The condition of the distribution transformer’s is evaluated in this article using real-time data from the transformer and specific sensors connected to Raspberry pi and artificial neural networks, are used to analyse the situation and make decisions regarding the health of the transformer. A model has been proposed for continuous monitoring consistent vigilance and swift actions against any faulty situations.