online voltage stability monitoring and prediction by using support vector machine considering overcurrent protection for transmission lines

In this paper, a novel method is proposed to monitor the power system voltage stability using support vector machine (svm) by implementing real-time data received from the wide area measurement system (wams). in this study, the effects of the protection schemes on the voltage magnitude of the buses are considered while they have not been investigated in previous researches. considering overcurrent protection for transmission lines not only resolves some drawbacks of the previous studies but also brings the case study system closer to the realities of actual systems. online monitoring of system stability is performed by prediction of the voltage stability index (vsi) and carried out by using support vector regression (svr). due to the direct effect of appropriate svr parameters on the prediction quality, the optimum value is chosen for learning machine hyperparameters using differential evolution (de) algorithm. the obtained simulation results demonstrate high accuracy, effectiveness, and optimal performance of the proposed technique in comparison with back-propagation neural network (bpnn) and adaptive neuro-fuzzy inference system (anfis) approaches. the presented method is carried out on the 39 bus new england system.