An advanced graph algorithm-based protection strategy for detecting kilometric and cross-country faults in DC microgrid

Abstract

Low voltage DC microgrids (LVDC) are on rise, because of increase in usage of electronics-based utility loads. However, the protection and safety aspects of these grids remain unresolved due to the fault current magnitude, unnecessary tripping, and blinding of protection. In this paper, an adaptive differential protection system incorporated with an advanced graph algorithm is proposed for DC microgrid. This graph algorithm is a combination of Fenwick tree and Bidirectional Dijkstra algorithm. Fenwick tree algorithm is used to determine the network configuration and Bidirectional Dijkstra algorithm is used to determine the least distance from the fault location to the nearest distributed generations. The developed protection scheme is applied to a 7 bus, 400 V DC microgrid setup using Real Time simulator (control hardware in loop) to detect and clear the faults such as kilometric faults, and cross-country faults. In terms of efficiency, the proposed algorithm demonstrates superiority over the conventional algorithm by detecting and rectifying faults in 384 μs.The results depict the robustness of the protection setup in clearing the faults rapidly with minimum network disconnection.

Keywords

• Bidirectional Dijkstra algorithm
• Distributed generations (DG)
• DC microgrids
• Fault detection
• Fenwick algorithm
• Protection controller (PC)