|
Abstract
|
Abstract: Early detection and localisation of winding faults within the transformer can help make preventive actions and
reduce potential damages to the transformer and the power system. This research work presents a new technique for
sensitive detection and localisation of shorted turns on the windings of power transformers throughout low-frequency
response measurements of the transformer windings. To this end, using genetic algorithm, the detailed model of the
damaged winding by the fault is estimated from the measured low-frequency response data up to 10 kHz. The fault is
localised along the winding by establishing the differences between the transfer functions of the healthy and faulty
state of the winding units using statistical indicators. The experimental measurements which were made on a test
transformer damaged by a low-level short-circuit fault proved that the newly developed method is sufficiently able and
sensitive to detect and localise failures of only few shorted turns on the transformer windings.
|