Artificial intelligence-based short-circuit fault identifier for MT-HVDC systems

The most convenient solution to link faraway significant renewable energy sources (RESs) is the voltage-source converter multi-terminal high-voltage DC systems (MT-HVDC). However, to maintain system stability and continuity of supply, a rigid and fast fault locating technique is required. This study proposes a novel inherent travelling waves based short-circuit DC fault identifier, which accurately identifies both of the fault location and faulty pole in multiple numbers of cables in MT-HVDC system using a single current sensor. Both of a discrete wavelet examiner and a fuzzy-neural pattern recogniser precisely spot the faulty line and fault location based on the mutual effects of short-circuit initiated travelling waves between lines belonging to the same loop. A software toolbox is structured to illustrate the adequacy of the proposed artificial intelligence technique. This method is valuable to MT-HVDC administration centres, particularly those concerned with long-distance RES.