Improved Median Mode Decomposition-Based Fault Detection Method in VSC Interfaced DC System

In the flexible dc transmission system based on voltage source converter (VSC)-HVdc, the current rises sharply after a fault occurs. Untimely, protection actions can affect the safe and stable operation of the transmission network as well as key equipment, so quickly and accurately identifying faults is a major challenge. In view of the traditional dc fault detection, fault identification is difficult, and the burden of detection algorithms is large; this article improves the traditional empirical mode decomposition (EMD) and proposes a fault transient detection method based on median mode decomposition (MMD). Compared with the traditional EMD, MMD only calculates in the time domain, which can separate the intrinsic mode function (IMF) with fewer iterations, reduce the computational burden of the detection algorithm in fault detection, and speed up the response speed; meanwhile, it introduces the instantaneous energy density level (IEDL) as the judgment basis for fault detection, which emphasizes the fault characteristics within a predetermined frequency range and suppresses the influence of steady-state components. Finally, simulations based on PSCAD/EMTDC have been carried out, and experimental verification by building a small multiterminal dc system demonstrates that the proposed method can complete the response within 2 ms within the fault occurrence. The suggested approach is well-suited for dc fault detection applications since it performs better than widely used transient analysis techniques in terms of response speed, sensitivity, noise resilience, and reliability.