Deep learning based digital twins for dynamic identification of high voltage grid connected converters – DELTWINCO

The production of electricity by renewable energy, connected to the grid with power electronic converters is increasing. Hence, the question of dynamic stability evaluation of electricity grid is becoming more and more complex due to power electronic converters. To cope with this situation, the Transmission System Operators needs to have a good dynamic knowledge of these new devices that are connected to the grid. One solution is to have a real-time evaluation of the power converter impedance. The proposed project will use some deep learning algorithms to develop a digital twin which mimics the dynamic behavior of the power converter in case of normal operation or large events such as short circuit. This digital twin, based on metamodel will allow the TSO to control the dynamic performance of the converters which are connected to the transmission grid.
Some experimentation will be developed on small-scale power electronic converters connected to a real-time simulator.