您的位置:首页  > 论文页面

基于CPU-GPU模型的暂态稳定异构并行算法

发表时间:2014-03-31  浏览量:2163  下载量:935
全部作者: 杜浩,严正,周挺辉
作者单位: 上海交通大学电子信息与电气工程学院,电力传输与功率变换控制教育部重点实验室
摘 要: 针对现代CPU和GPU多核心的特点,对电力系统暂态稳定分析及其并行算法进行研究。结合电力网络矩阵稀疏化特征与计算统一设备架构(compute unified device architecture,CUDA)技术,提出基于CPU网络矩阵划分及运算调度和GPU细粒度并行的电力系统暂态稳定分析的异构并行算法。利用多个案例对方案进行测试,结果表明:利用GPU及CUDA技术能够获得良好的加速比效果,且存在巨大的提升空间。
关 键 词: 电力系统及其自动化;暂态稳定;异构并行;计算统一设备架构
Title: Research of CPU-GPU based transient stability heterogeneous parallel computing
Author: DU Hao, YAN Zheng, ZHOU Tinghui
Organization: Key Laboratory of Control of Power Transmission and Conversion, Ministry of Education, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University
Abstract: For modern multi-core CPU and GPU features, a research is made on power system transient stability analysis and its parallel computing. Considering the sparse feature of power network grid and compute unified device architecture (CUDA) technology, a method of power system transient stability heterogeneous parallel computing based on CPU network grid partition and calculation dispatch, together with GPU fine-grained parallelism is proposed. This study tests the solution with a series of cases. Test results show that programs can gain a favorable speed-up ratio with the GPU and the CUDA technology, and there is a huge room for improvement.
Key words: electric power system and automation; transient stability; heterogeneous parallel computing; compute unified device architecture
发表期数: 2014年3月第6期
引用格式: 杜浩,严正,周挺辉. 基于CPU-GPU模型的暂态稳定异构并行算法[J]. 中国科技论文在线精品论文,2014,7(6):536-543.
 
0 评论数 0
暂无评论
友情链接