Voltage stability security margin assessment via artificial neural networks

This paper presents an alternative approach to estimate voltage stability margins (VSM), both under normal operating conditions and contingencies, based on the use of one only multilayer artificial neural network (ANN). The ANN is trained using measurable quantities obtained directly from the system, or easily calculated ones obtained from load forecasted data. An input data set reduction procedure is also included, in order to guarantee the efficiency of the proposed method. Simulation results for the IEEE systems are shown to demonstrate the effectiveness of the proposed ANN.