Incipient Fault Localization and Classification for Transmission Lines using Neural Intelligent Technique
The problem of incipient fault localization and classification in power transmission lines is an emerging area of power system research that seeks to determine the likelihood or probability of fault just before its occurrence. This involves the determination of power line fault signatures and online characterization of line parameters. This research paper applies a simulations and data driven based approach emphasizing resonance theory of transmission lines and neural intelligence for effective fault location determination and incipient fault prediction in transmission lines. Simulations considering the NeuroAMI predictor for the PSD signals showed that apart from peaks of about 25V/k-Hz, 27V/k-Hz and 35V/k-Hz, the proposed neural predictor fault location estimates closely matched the expected fault locations. Considering the data-driven approach based on a public dataset, the proposed NeuroAMi technique showed superior RMSE values over the conventional BP-FFANN.