Online intelligent technique for preventing relay maloperation under stressed conditions

Various power system blackouts have been caused due to the maloperation of distance relays during stressed conditions like power swing and voltage instability. Thus differentiating fault from stressed conditions and making the protection scheme intelligent enough to stop the relay maloperations has become very important. There are a few computational intelligent techniques proposed in the literature for preventing relay maloperations. However with the increase in size and complexity of the power systems there have been situations during which there is change in network topology or system parameters. An online intelligent technique: online sequential extreme learning machine (OSELM) has been suggested in this paper which under such real time situations successfully furnishes accurate results. This online computational intelligence technique based on synchronized wide area measurements has been implemented to develop a classifier that differentiates fault from power swing and voltage instability. Potential of this online tool in preventing relay maloperations has been validated by comparing it with other offline intelligent techniques during real time scenario.