An Intelligent Faulty Line Identification Scheme for Micro-grids

Abstract

The available micro-grid protection schemes in the literature use the features extracted from two ends of the feeder for fault detection but have failed to report the simultaneous status of the adjacent feeder. This paper presents an intelligent faulty line identification scheme using k-nearest neighbor (k-NN) to identify the appropriate faulty feeder. Time-synchronized measurements are retrieved and pre-processed using discrete Fourier transform (DFT). Unlike any other time–frequency domain methods which have computation burden, here the simple statistical features computed from the DFT coefficients are used as input features. Separate k-NN modules have been designed for fault detection and faulty line identification to provide complete protection of the micro-grid. An extensive study has been performed on a standard micro-grid by varying different operating conditions in all micro-grid configurations. Results validate the efficacy of the proposed methodology.