Power System Security Margin Prediction Using Radial Basis Function Networks
This paper presents a method to predict the postcontingency security margin using radial basis function (RBF) networks. A genetic algorithm-based feature selection tool is developed to obtain the most predictive attributes for use in RBF networks. The proposed method is applied to a thermal overload problem for demonstration. Simulation results show that the proposed method gives satisfactory results and the running time decreases by a factor of 10 compared with using multilayer perceptrons.