Health Index prediction using Artificial Neural Network (ANN) on Historical Data of Power Transformer

Abstract

Power transformer is an important equipment in the electric power system. The power transformer has the main task of changing the voltage, transmitting electricity, and distributing electricity. Disruption or failure of the transformer can result in asset fires and power outages. Transformer power failure can result in social and economic losses. The right maintenance strategy can reduce the risk of transformer failure and optimize operational costs and maintenance costs. The Health Index is used to provide an overall assessment of the condition of the power transformer, assess the reliability of the power transformer, and the strategy for maintaining the power transformer. In addition to durability, the health index of the transformer can be assessed from the life expectancy of the transformer. Health Index values can be obtained from expert judgment, calculations, and prediction methods using Artificial Intelligence. This paper discusses the implementation of Artificial Neural Network (ANN) as one of the Artificial Intelligence (AI) algorithms to predict the condition of the transformer health index. The result is compared to the calculated HI, then validated by 79 transformers that have been comprehensively assessed by the expert.