Paper Details

This paper gives a rough idea about the use of artificial neural network to the protection of power transformer. The protective system include devices that recognizes the existence of a fault, indicates its location and class, detect some other abnormal fault like operating condition and start the nascent steps of opening of circuit breaker to disconnect the faulty equipment of power system. The ANNs in these existing studies are specific to particular transformer systems, and would need to be retrained again for other systems. In case of incipient fault protection using Dissolved Gas Analysis (DGA), initially a single Artificial Neural Network (ANN) with three layer architecture is developed that have the best performance for individual fault diagnosis. But, when a single ANN is used for individual fault diagnosis, the accuracy and training speed are low. Also sometimes data availability may be insufficient and inconsistent for ANN training. Therefore, a combined ANN and Expert System #40; ANNEPS#41; tool is developed for power transformer incipient fault diagnosis.

Expert System, Fuzzy, Neural, Transmission Line, DGA