Application of neural networks to power system security: technology and trends

This paper presents an overview of the application of artificial neural networks (NN) to power system security assessment. It is noted that although the majority of NN architectures used is the multilayered perceptron, some work has been done to use the Hopfield and the Kohonen networks. In either case, the present applications are illustrated using small power systems, and the key issues are the selection of the input data, training set and the evaluation of the NN design in terms of its accuracy in predicting the security of the power system. Most of the discussion in this paper is concerned with the latter issue since it has not been addressed extensively in the literature.<
>