High Impedance Fault Detection based on Mathematical Morphology for Radial Distribution Network

This manuscript presents the use of the mathematical morphological (MM) on fault detection and pattern recognition. On this basis, a new approach for high impedance fault detection using MM is introduced, which employs the morphological gradient to obtain faulty feature indices from statistical properties of dilation and erosion operators to establish feature index and finally, this vital feature index is compared with a pre-defined threshold to complete the HIF detection task. To validate the performance of the proposed approach, several NON-HIF events (capacitor switching, linear and non-linear load switching, motor starting and low-impedance faults) along with the HIFs are simulated and tested through the proposed detection algorithm. Results clearly show that the proposed scheme takes single cycle for HIF detection after the initiation of fault and therefore, the stated method has been appropriate for HIF detection in electric distribution systems including the widespread diverse operative environment.