Fault detection methods for medium-voltage flexible DC power supply systems under active current limiting control at MMC converter stations

In medium-voltage flexible DC power supply systems, conventional protection of Modular Multilevel Converters (MMC) typically relies on MMC blocking, which may result in converter station shutdown. Under active current-limiting control, the fault current is suppressed by reducing the number of inserted submodules during a fault. This strategy enables DC-side faults to be detected before the MMC converter station is blocked or shut down, thereby providing sufficient transient time for subsequent fault detection and selective isolation. However, the suppression of fault currents via active current-limiting control may reduce the accuracy of conventional. protection methods in fault detection. To address this issue, this paper proposes an Logistic-Gauss Circle- Feature Mode Decomposition(LGC-FMD)-based fault detection method for medium-voltage flexible DC power supply systems that incorporates converter-station current-limiting control. By combining this method with a Joint Judgment Criteria of IMF based on Approximate Entropy(AE) and the Maximal Information Coefficient(MIC), Despite the fault current suppression provided by active current-limiting control at the converter station, faults can still be accurately detected. This method is of practical importance for improving the stability and reliability of medium-voltage flexible DC power supply systems.