Ultra High-Speed Fault Diagnosis Scheme for DC Distribution Systems Based on Discrete Median Filter and Mathematical Morphology

Abstract

In the modern world, the growing presence of renewable energy power near the consumer end makes DC distribution systems more attractive than traditional AC ones. However, designing fault diagnosis schemes for the DC distribution system is a complicated problem due to noisy measurements and rapidly rising DC-fault currents. This article proposes an Ultra High-speed Fault Diagnosis scheme using a Discrete Median Filter and Mathematical Morphology Algorithm. In the first stage, the acquired current signal from a corresponding faulty bus is preprocessed using a Discrete Median Filter for state estimation and noise reduction. In the second stage, the proposed Scheme computes the DC Residual by deploying the Mathematical Morphology Algorithm on the DMF estimated current signal. The DC Residual represents the mathematical computed difference between the filtered output of the Mathematical Morphology Algorithm and the Discrete Median Filter-estimated current signal. Then, the proposed method cross-matches variations in the computed DC Residual with pre-defined threshold settings to detect faults successfully and promptly. In the third stage, the Mathematical Morphology Algorithm-based Energy is computed for fault classification and Section identification. The polarity of Mathematical Morphology Algorithm-based Energy is used to categorize and locate all types of DC faults in the proposed scheme. The suggested method is tested on a DC distribution test bed via MATLAB/Simulink 2022b. The results demonstrate that the suggested scheme successfully identifies all types of DC faults in less than 2.5 msec, with 99% accuracy.

Keywords

• Discrete median filter
• fault detection
• mathematical morphology
• renewable energy-based DC distribution network
• smart grids