A new scheme for single-pole autoreclosure based on recurrent ANNs

This paper outlines new techniques for distinguishing between transient and permanent faults and for identifying the secondary arc extinction time for faults encountered on typical 400 kV uncompensated transmission lines, employing single-phase autoreclosure schemes. Presented outcomes show that ANNs consisting of very few neurons equipped with feedback are able to successfully process post-initial circuit breaker (CB) opening voltage waveforms for a decision making purpose. Moreover, in comparison with hitherto employed ANNs the herein presented scheme does not require any sophisticated pre-processing of the input signal, e.g. spectral analyses, feature extraction.