A GPU-based parallel simulation platform for large-scale wind farm integration

Due to the energy crisis of the next fifty years, an increasing number of large-scale wind farms are integrating into power systems. Parallel simulation platforms are indispensable to evaluate the effects of wind farm integration. A GPU-based parallel simulation platform is constructed in this paper to test the feasibility of simulating integration systems on GPU. By adopting previous research work such as averaged model and improved EMTP algorithm, models and algorithm of the platform are firstly introduced. Then the hardware and software architecture of the platform are put forward, which combine CPU configuration and GPU computing. Finally, a wind farm with different numbers of DFIGs are used to test the performance of the platform. Results show that the platform has excellent accuracy, efficiency and scalability. The work of this paper proves the feasibility to simulate wind farm integrating dynamics adopting GPU and its potential to reach tremendous efficiency and scalability.