A multiple model approach to fault detection and localization in multi-terminal HVDC systems

This paper discusses a novel protection technique for DC line faults in multi-terminal HVDC (MTDC) systems. Hereby a bank of Kalman filters ranks the fit of a discrete set of line fault models to measured data. The best fitting model indicates the fault location and fault type. The suitability of the technique is discussed for fast on-line fault localization and exact off-line fault localization and fault identification. Experimental results are shown to validate the approach based on data from a high fidelity simulation of a 3-terminal HVDC system.