Two-Stage Parallel Waveform Relaxation Method for Large-Scale Power System Transient Stability Simulation

Abstract

This work presents a parallel transient stability simulation based on waveform relaxation for large-scale power systems. A power system described by large-scale differential-algebraic equations (DAE) is decomposed into several subsystems by a partitioning algorithm based on epsilon decomposition. The method adopts preconditioning and waveform prediction to accelerate convergence of the system. Moreover, a two-stage parallel strategy based on OpenMP further improves parallel efficiency. All subsystems are solved using a very dishonest Newton (VDHN) method combined with an iteration method based on Adomian decomposition independently after the state variables are discretized by implicit-trapezoidal rule. Finally, two large-scale test cases with detailed dynamic models verify the proposed algorithm. The computational results show that the proposed waveform relaxation method achieves high parallel efficiency.