文章摘要
周益清,骆文于,吴双林.三维声传播模型BELLHOP3D的MPI并行优化*[J].,2023,42(1):93-99
三维声传播模型BELLHOP3D的MPI并行优化*
MPI parallel optimization of 3D sound propagation model BELLHOP3D
投稿时间:2021-12-01  修订日期:2023-01-03
中文摘要:
      近些年,随着我国对海洋探索的不断深入,对复杂环境中声场的快速预报提出了越来越高的需求。BELLHOP3D是一种基于射线法的三维声传播计算模型,在海洋声学中应用十分广泛。BELLHOP3D的计算效率比其他常用模型高,但是仍然有非常大的提升空间。本文使用MPI(消息传递接口)对BELLHOP3D进行粗粒度的并行优化,并行后的程序计算结果稳定可靠,并行效率高,更适合在实际应用中实现快速的声场预报。本文首先介绍射线法的原理,然后介绍BELLHOP3D的串行算法和并行优化算法,之后通过数值实验验证并行程序的计算准确性并测试它的并行计算性能。计算结果表明使用MPI对BELLHOP3D进行并行优化,计算精度非常高,而且能大幅节约计算时间。并行BELLHOP3D程序可以在https://github.com/nj-zyq/BELLHOP3D_MPI.git下载。
英文摘要:
      In recent years, with the development of marine science in China, there is a more and more demand for the rapid prediction of sound field in complex environment. BELLHOP3D is a sound propagation calculation model based on ray method, which is widely used in ocean acoustics. BELLHOP3D has higher computational efficiency than other commonly used models, but it is still possible to improve its efficiency. In this paper, the coarse-grained parallel optimization of BELLHOP3D is carried out by using MPI (Message Passing Interface). The results calculated by parallel program are stable and reliable, and the parallel efficiency is high, which is more suitable for realizing fast sound field prediction in practical application. This paper first introduces the theory of ray method, then introduces the serial algorithm and parallel algorithm of BELLHOP3D, and then verifies the computational accuracy of parallel program through numerical experiments and finally tests its parallel optimization computing performance. The calculation results show that the parallel optimization of BELLHOP3D using MPI has very high calculation accuracy and can greatly save calculation time. Code is available at https://github.com/nj-zyq/BELLHOP3D_MPI.git.
DOI:10.11684/j.issn.1000-310X.2023.01.012
中文关键词: 海洋声学  水下声传播  射线法  BELLHOP3D  MPI并行
英文关键词: ocean acoustics  underwater sound propagation  ray method  BELLHOP3D  MPI
基金项目:国家自然科学基金项目(面上项目,重点项目,重大项目)
作者单位E-mail
周益清* 中国科学院声学所 zhouyiqing18@mails.ucas.ac.cn 
骆文于 中国科学院声学所 lwy@mail.ioa.ac.cn 
吴双林 中国科学院声学研究所 声场声信息国家重点实验室  
摘要点击次数: 616
全文下载次数: 623
查看全文   查看/发表评论  下载PDF阅读器
关闭