A method for fault location in distribution systems

This paper describes a method for fault location in power distribution systems. The method is fundamentally based on artificial neural networks. It also employs transmission line theory and the resistive nature of the fault impedance. The method is applicable to feeders containing lateral derivations (single or three phase) and nonhomogeneous characteristic. Phase current and voltage fundamental frequency phasors at the origin of the feeder are needed as input data for a diagnostic. Studies regarding the influence of errors on the system modeling on the results are also presented.