Artificial neural network approach to single-ended fault locatorfor transmission lines

摘要

This paper describes the application of an artificial neuralnnetwork-based algorithm to the single-ended fault location ofntransmission lines using voltage and current data. From the faultnlocation equations, similar to the conventional approach, this methodnselects phasors of prefault and superimposed voltages and currents fromnall phases of the transmission line as inputs of the artificial neuralnnetwork. The outputs of the neural network are the fault position andnthe fault resistance. With its function approximation ability, thenneural network is trained to map the nonlinear relationship existing innthe fault location equations with the distributed parameter line model.nIt can get both fast speed and high accuracy. The influence of thenremote-end infeed on neural network structure is studied. A comparisonnwith the conventional method has been done. It is shown that the neuralnnetwork-based method can adapt itself to big variations of sourcenimpedances at the remote terminal. Finally, when the remote sourcenimpedances vary in small ranges, the structure of artificial neuralnnetwork has been optimized by the pruning method