Abstract:
The ongoing proliferation of distributed energy resources and power electronic converters at distribution level of modern power systems causes control and protection issues and therefore conducting real-time simulation studies is an indispensable and essential step to verify performance of control devices and protection relays prior to installation. Conducting these studies for modern distribution systems faces computational challenges due to (i) large number of nodes, components and therefore size of differential algebraic equations, (ii) short transmission lines, (iii) large number of repetitive-switching devices, i.e., power electronic converters and (iv) small real-time simulation time-step requirement due to high switching frequency of these converters.
FPGA features parallel and reconfigurable computing that can be utilized to design an application-specific hardware to overcome the computational challenges associated with real-time simulation of modern distribution systems. However, the existing power system component models are not suitable for FPGA-based real-time simulation due to inherent differences between CPU- and FPGA-based computing. We develop (i) new component models, tailored to exploit parallel processing capabilities of FPGAs, and (ii) an FPGA-based real-time simulator suitable for modern distribution systems. This simulator is connected to the industry standard real-time simulator, i.e., RTDS, to form a versatile real-time simulation platform in which the distribution system is simulated in the FPGA-based simulator and the rest of the power system is simulated in the RTDS. This platform is applicable to real-time simulation studies of microgrids, virtual power plants, wind farms, solar farms and also all-electric ship and aircraft power systems. In this presentation the FPGA-based power system component models and the FPGA-based real-time simulator are introduced and their capabilities are explained.
Ramin Mirzahosseini ~ University of Toronto