Applications of neural-net computing to transient stability assessment and enhancement of electric power systems

This paper describes some application of artificial neural networks in electric power systems. The concept of adaptive pattern recognition and neural networks in the process of recognizing and classifying patterns is explained. The Generalized Delta Rule and the Functional Link Net architectures of neural nets and the unsupervised/supervised learning concept are discussed. Generalization capabilities of neural nets are illustrated in applications on unstable machine identification, calculation of voltage dips and calculation of generation shedding requirements.