欢迎光临《计算力学学报》官方网站!
基于CUDA的有限元矩阵并行装配算法研究
Research on Parallel Assembly Algorithms of Finite Element Matrices Based on CUDA
投稿时间:2019-05-14  修订日期:2019-07-23
DOI:
中文关键词:  有限元  数值计算  特征矩阵组装  CUDA
英文关键词:Finite element method  Numerical Integration  Stiffness matrix assembly  CUDA
基金项目:国防基础科研项目
作者单位E-mail
胡斌星 西北工业大学 376898978@qq.com 
李新国 西北工业大学  
摘要点击次数: 691
全文下载次数: 0
中文摘要:
      构建航天飞行器的结构有限元模型是准确模拟飞行仿真、完成飞行器在轨飞行阶段结构故障监测和诊断的基础。本文采用细长体飞行器简化梁模型,提出新的基于CUDA(Compute Unified Device Architecture)的有限元单元刚度矩阵生成和总刚度矩阵组装算法。依据梁单元单元矩阵的对称性,结合GPU硬件架构提出并行生成算法并进行改进。为有效减少装配时间,在装配过程中采用著名的着色算法,提出了基于GPU(Graphics Processing Unit)共享内存的非零项组装策略,通过在不同计算平台下算例对比,验证了新算法的快速性。数值算例表明,本文算法的求解效率较高,针对一定计算规模内的模型可满足快速计算与诊断的实时性要求。
英文摘要:
      Constructing the finite element model of spacecraft structure is the basis of flight simulation, structural fault monitoring and diagnosis during the on-orbit flight phase. A new stiffness matrix generation and stiffness matrix assembly algorithm for finite element method based on CUDA is proposed based on the simplified beam model of slender body aircraft. According to the symmetry of beam element matrix, the parallel algorithm is proposed and improved by combining with GPU hardware architecture. In order to reduce the assembly time effectively, the famous coloring algorithm was used in the assembly process, and a non-zero assembly strategy based on GPU Shared memory was proposed. Numerical examples show that the algorithm presented in this paper is efficient, The algorithm satisfies the real-time requirement of fast calculation and diagnosis for the model in a certain calculation scale.
  查看/发表评论  下载PDF阅读器
您是第5906818位访问者
版权所有:《计算力学学报》编辑部
本系统由 北京勤云科技发展有限公司设计