Power system security boundary visualization using neural networks

This paper presents an automatic security boundary visualization procedure using neural networks. A systematic method for data generation is developed to generate a database for neural network training. Neural networks are us ed to map the relationship between the precontingency operating parameters and the postcontingency performance measure. Genetic algorithms are used to select best subset of precontingency operating parameters to be used as ne ural network inputs. A visualization algorithm is developed to draw the bou ndary. The boundary for a sample system is given. (C) 1998 Elsevier Science B.V. All rights reserved.