Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
The development of modern and future power systems is associated with the definition of new approaches for their simulation, control, and protection. To give an example, the increasing connection of massive renewable energy conversion systems is justifying the integration of DC infrastructures (eventually, multi-terminal HVDC) in the current AC power grids. Furthermore, the existing passive distribution networks are evolving by integration of decentralized and intermittent generation units which results in Active Distribution Networks (ADNs). As a consequence, complex power system topologies are emerging requiring adequate simulation tools capable to reproduce, possibly in real-time, their dynamic behavior. In this context, future operation/protection practices of power networks might rely on the availability of chip-scale real-time simulators (RTS) that will enable the implementation of efficient protection/fault location processes that, in principle, should be capable to comply with the restrictive constraints associated with these complex systems. Within this context, the work presented in the thesis contributes to the integration of new concepts of the fault location in AC/DC systems that can be deployed in chip-scale real-time simulation hardware represented by Field Programmable Gate Arrays (FPGAs). The development of the proposed fault location platform is done in two steps. First, an original fault location method based on the Electromagnetic Time Reversal (EMTR) theory is proposed. The proposed method is validated for the case of various power networks topologies and its performance is assessed. Compared to the existing fault location methods, the proposed approach is suitably applicable to different topologies including MTDCs and ADNs. Next, a new automated FPGA-based solver for RTS is proposed. The developed FPGA-RTS uses a specific automated procedure to couple the simulation platform with an offline simulation environment (EMTR-RV) without the need for Hardware Description Language (HDL). It is able to simulate both power electronics converters and power system grids and thanks to the use of particular parallel computational algorithms, it can accurately simulate, in real-time, Electromagnetic Transient (EMT) phenomena taking place in power converters and travelling wave propagation along multi-conductor transmission lines within very small simulation time steps (in the order of some hundreds of nanoseconds). To overcome the limitations associated with the Fixed Admittance Matrix Nodal Method (FAMNM), a method to assess the optimal value of the parameter of the Associated Discrete Circuit (ADC) switch model used by FAMNM is proposed. Finally, a specific application of the developed FPGA-RTS is explored for the development of a fault location platform by leveraging the EMTR theory. To this end, the proposed EMTR-based fault location method is integrated with the FPGA-RTS to develop an efficient fault location platform. Thanks to the fast EMT simulation capability of the FPGA-RTS, the developed fault location platform is able to estimate the accurate fault location within very short time scales. Moreover, the developed platform is compatible with the constraints characterizing complex topologies such as MTDC networks (e.g., the ultra-fast operation of the protection systems). The developed fault location platform is validated by making reference to an MTDC grid and an ADN, and it is shown to exhibit remarkable fault location accuracy as well as robustness against uncertainties such as fault type, the presence of noise, measurement systems delay, and fault impedance.