Power system transient stability assessment based on online sequential extreme learning machine

Recently, pattern recognition-based transient stability assessment methods have shown much potential for on-line transient stability assessment (TSA) of power systems. However, the current models usually suffer from excessive training time and parameter tuning difficulties, leading to inefficiency for online model updating. Considering the possible real-time information provided by phasor measurement units, a new TSA method based on online sequential extreme learning machine is proposed in this paper. The presented method can efficiently update the trained model on-line by partial training on the new data to reduce the model updating time whenever a new special case occurs. The effectiveness of the proposed method is validated by the simulation results on the New England 39-bus test system.