Prony preprocessing of radial basis function network for the fault localization on transmission lines

This paper presents the application of Prony analysis and radial basis function technique in estimating the fault location on transmission lines for a two area test system. Prony method is used for data pre-processing and this data are further used for training and testing radial basis function (RBF) network. Prony analysis is used to extract the modal contents and inter-area oscillation symptoms of the fault. The radial basis function network is first trained using the modal information (frequency, damping, energy) provided by the Prony analysis. It is found that the proposed method gives satisfactory results and will be useful for estimating the fault location.