A New Method to Detect DMR Faults During Balanced Condition for a VSC Bipole Onshore-Offshore HVDC Network

The bipole High-Voltage Direct Current (HVDC) systems can transmit high power rating when compared with monopole HVDC link. The flexibility and reliability that is provided by the bipole configuration make it an attractive solution for power transmission from offshore wind farms to onshore grid. These bipole HVDCs have dedicated metallic return (DMR) as a return path and ground references. The detection of fault in the DMR cable of a bipole HVDC is a challenging task. Traditional over-current and differential protection fail to detect the faults in the network. In this paper, the moving window-based Pearson correlation coefficient (MWPCC), variational mode decomposition (VMD) and bagging ensemble classifier (BEC) based two fault detection algorithms are proposed. The first two advanced signal processing methods are utilized to extract data for the classifier. The effectiveness of the proposed detection technique is demonstrated by applying different fault scenarios. The time domain simulations are performed using a full-scale industrial model, and the fault detection algorithms have achieved very high accuracy to detect the fault.