IPSO Based Fault Analysis in Power Transformer

A power transformer is a mechanical device that converts the voltage of a circuit to another without altering the frequency. A power transformer is the most expensive device in the electrical systems. The transformer failure would result in huge economic loss and unexpected outage of power system; hence a maintenance mechanism is essential to prevent the transformers from failures. Components may fail due to poor maintenance, poor operation, poor protection, undetected faults, severe lightening, short circuits, etc. The replacement of the faulty component is a time consuming and an expensive process. The average lifetime of a transformer is more than 30 years. During this period, the transformers demand proper maintenance to increase their life expectancy. Faults in any component of the transformer would result in heavy economical loss, an efficient fault diagnostic techniques should be incorporated to prevent the loss. In this research, an efficient optimization technique named, IPSO-RBF is proposed to diagnose and classify the fault that occurs in the power transformer. Primary RBF is used to extract the features from the DGA dataset; these features are the input data for performing fault analysis in IPSO-RBF. The DGA dataset for proposed system are taken from diagnostic gas in oil of 500 KV main transformers of Pingguo Substation in South China Electric Power Company. The comparative analysis is made in order to evaluate the performance of the classifiers PSO-RBF and IPSO-RBF in terms of classification accuracy and time. Finally, the result shows that the proposed IPSO-RBF has greater precision rate and computational time for fault analysis using DGA dataset.