Application of GPU-based Block Reduction Algorithm in Power System Small-signal Stability Analysis

ZHANGYifei,YAN Zheng,ZHAO Wenkai,CAO Lu,LI Jianhua
DOI: https://doi.org/10.7500/AEPS20150126012
2015-01-01
Abstract:To enhance the computational efficiency of complete eigenvalue analysis in power system small-signal stability analysis,the parallelization of upper Hessenberg reduction algorithm in the QR method is studied.A block reduction algorithm is utilized to integrate the floating-point operations into high-level basic linear algebraic subprograms (BLAS).The block reduction algorithm is parallelized on hybrid CPU/GPU (graphic processing unit) system and applied to the complete eigenvalue analysis of large-scale power system small-signal stability analysis.Simulation results show that,compared with multi-core CPU parallelization,the GPU-based block upper Hessenberg reduction algorithm is able to obtain a speed-up ratio up to 5 times the original.The overall computing speed of the complete eigenvalue analysis,including the method proposed, has achieved remarkable acceleration improvement.The applicability of the QR method to large-scale power system simulation analysis is increased.
What problem does this paper attempt to address?