Transient Stability Prediction using Artificial Neural Networks and Synchronized Measurements

In this paper an post-fault transient stability assessment (TSA) method in terms of transient stability margin (TSM) prediction using artificial neural networks (ANN) and synchronized or PMU measurements is proposed. A post-fault multi-machine system is converted into a suitable OMIB using Single Machine Equivalent (SIME) concept. By using SIME Pa-? trajectory, a normalized transient stability margin is calculated. By using pre and during fault synchrophasor measurements as input ANN model is trained to predict normalized stability margin. Using the synchronized measurements available at generator buses and the trained ANN model, post-fault TSA is carried out in terms of TSM prediction. If the predicted margin is negative then the post-fault system is declared unstable and if the predicted margin is positive then the system is declared stable. The proposed assessment method is implemented using New England 39 bus test system. The results are compared with time domain simulations.