High Speed Protection of Medium Voltage DC Distribution System using Modified Mathematical Morphology /

Abstract

One of the Challenging issues in the realization of medium voltage DC distribution system is high speed protection scheme design against quickly rising short circuit fault-DC current. This thesis proposes a high-speed protection scheme based on the processing of current signal through mathematical morphology. The proposed scheme uses a modified MM-based filter to extract the transient information in the current signal rapidly and distinctly in the form of positive/negative polarity. The polarity detection of any disturbance in the analyzed current signal indicates the occurrence of fault in the system. The fault occurrence confirmation triggers the fault location and classification algorithms of the proposed scheme in order to locate the faulty distribution line section and to classify the fault. The fault location and classification algorithms are based on the detected polarities of the current signal. The protection scheme requires low bandwidth communication. The proposed protection scheme is evaluated by using a ±2.5 kv radial distribution network containing 4 feeders under different fault conditions in MATLAB/SIMULINK software. Simulation results verify that the proposed scheme can isolate the faulty portion within few milliseconds after fault inception under variety of fault cases.