Signal Processing Based High Speed Protection of MVDC Distribution System Using Mathematical Morphology

Abstract

The design of high speed protection schemes against rapidly increasing short circuit fault-DC current is one of the difficult problems in the realisation of medium voltage DC (MVDC) distribution systems. The high-speed protection method proposed in this research uses mathematical morphology to process energy signals . The suggested method makes use of a modified MM-based filter to quickly and clearly extract the transient information from the energy signal in the form of positive/negative polarity. Any abnormality in the examined energy signal's polarity that can be detected indicates a systemic malfunction. The proposed scheme's fault location and classification algorithms are triggered by the confirmation of a fault in order to find the faulty distribution line section (DLS) and categorise the issue. The energy signal's identified polarities serve as the foundation for the algorithms used to locate and categorise faults. Low bandwidth communication is necessary for the protection strategy. Using a 2.5 kV radial distribution network with four feeders and several fault scenarios, the proposed protection strategy is assessed using the MATLAB/SIMULINK software. The suggested scheme's ability to isolate the defective section within a few milliseconds of fault initiation in a range of failure scenarios is supported by simulation findings.