Power System Transient Stability Prediction Using Convolution Neural Network and Saliency map

Transient stability of a power system is one of the most important indices in power system operation, as it represents the criterion for determining whether the power system can continue to operate without losing synchronization in the event of accidents or faults occurring in the system. This paper proposes a convolutional neural network (CNN) model combined with a saliency map to predict the transient stability of a power system. The CNN model learns power system transient stability using the saliency map as input. The proposed method shows higher accuracy than the existing method, which uses RGB images as input. The accuracy of the proposed method is verified using simulation data obtained from the IEEE 39-bus system through Matlab/Simulink.