Generalization of transient stability solution using neural network theory

Abstract

This paper presents a generalized online transient stability solution technique using the backpropagation method of the neural network theory. The proposed solution technique can be used for general generator models, including controllers such as AVRs and governors, that have been difficult for the energy function method to handle. The proposed method also considers alterations of network configurations and changes of the number of operating generators. Further, the relationship between input data and effects on the results of the neural network is considered.