New approach to transient stability prediction of power systems in wide area measurement systems based on multiple-criteria decision making theory

To prevent power system blackouts following a large disturbance such as a fault, the speed and precision of the stability status prediction are two important factors for the remedial action schemes. A new approach for the transient stability prediction is presented employing the multi-criteria decision making (MCD) theory to consider both factors simultaneously in this study. It is shown that they depend on the prediction time, therefore solving the MCD problem results in finding the best time for the stability prediction. As the criteria quantifying is necessary in the MCD methods, in this study a new index is proposed to quantify the prediction speed. Furthermore, the support vector machine (SVM) classifier is employed as a transient stability predictor. In addition, the best features are selected and incorporated as the inputs of the SVM classifier according to the features selection methods. Finally, all proposed schemes are implemented on the IEEE 39-Bus test system. The obtained results show the prediction of stability at the optimum time causing the instability prevention with the less required remedial action.