计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (5): 101-108.doi: 10.11896/jsjkx.241100169

• 高性能计算 • 上一篇    下一篇

基于局部加密网格的LBM程序负载均衡方法研究

黄晨希, 李家辉, 颜辉, 钟英, 卢宇彤   

  1. 中山大学计算机学院 广州 510006
    国家超级计算广州中心 广州 510006
  • 收稿日期:2024-11-27 修回日期:2025-02-28 出版日期:2025-05-15 发布日期:2025-05-12
  • 通讯作者: 李家辉(jiahui.li@nscc-gz.cn)
  • 作者简介:(chenxi.huang@nscc-gz.cn)

Investigation on Load Balancing Strategies for Lattice Boltzmann Method with Local Grid Refinement

HUANG Chenxi, LI Jiahui, YAN Hui, ZHONG Ying, LU Yutong   

  1. School of Computer Science and Engineering,Sun Yat-Sen University,Guangzhou 510006,China
    National Supercomputing Center in Guangzhou,Guangzhou 510006,China
  • Received:2024-11-27 Revised:2025-02-28 Online:2025-05-15 Published:2025-05-12
  • About author:HUANG Chenxi,born in 1994,postgraduate.His main research interests include computational fluid dynamics and high-performance computing.
    LI Jiahui,born in 1984,Ph.D.His main research interests include computational fluid dynamics and high-performance computing.

PDF (PC)

53

摘要： 基于局部加密网格的格子玻尔兹曼方法(LBM),在大规模的非稳态计算流体力学问题中有着越来越广泛的使用。局部网格加密方法虽然可以有效减少计算量,但是也带来了严重的负载均衡问题。尤其是在大规模并行计算中,负载均衡算法的选择对整体的计算效率有重要的影响。使用自研的LBM程序,简单对比了开源框架Palabos的实现结果,然后探讨了基于局部加密网格格子玻尔兹曼方法的静态负载均衡算法。程序中分别基于全局负载和分层负载,使用贪心算法和空间填充曲线算法实现了多种负载均衡策略,提出了计算节点负载均衡优化,并使用小球绕流算例和DrivAer算例进行测试和比较。根据计算结果可知,基于分层负载策略的性能比基于全局负载策略的性能要好;同时,贪心算法和空间填充曲线算法对比各有优劣,前者有较好的可扩展性,而后者在进程数较少时效率较高;最后,基于分层负载,结合贪心算法和空间填充曲线算法实现了混合算法,在进程数较多时获得了最佳性能。

关键词: 格子玻尔兹曼方法, 负载均衡算法, 局部加密网格, 并行计算, 节点负载均衡

Abstract: The Lattice Boltzmann method(LBM) with local mesh refinement is widely used in large-scale unsteady computational fluid dynamics problems.Although the local mesh refinement method can effectively reduce the computational workload,it also brings serious load balancing issues.Especially in large-scale parallel computing,the choice of load balancing algorithms has a significant impact on the overall computational efficiency.This paper used a self-developed LBM program to compare the implementations of Palabos and to explore the static load balancing algorithm based on the Lattice Boltzmann method with local mesh refinement.This paper implementes various load balancing strategies based on global load and stratified load,uses greedy algorithms and space-filling curve algorithms,and proposes optimization of the load balancing of computing nodes.It uses the flow around a sphere and the DrivAer case as test cases for testing and comparison.Firstly,the results show the performance based on the stratified load strategy is better than that based on the global load strategy.At the same time,a comparative analysis reveals that the greedy algorithm demostrates superior scalability,whereas the space-filling curve algorithmexhibits higher efficiency when operating with a limited number of process.Finally,based on layered load,a hybrid algorithm is implemented by combining greedy algorithm and space filling curve algorithm,which achieves the best performance when the number of processes is larger.

Key words: Lattice Boltzmann method, Load balancing, Local grid refinement, Parallel computing, Node-wise load balancing

中图分类号: 

  • TP319
[1]HE Y L,WANG Y,LI Q.Lattice Boltzmann method:Theory and applications [M].Beijing:Science Press,2009.
[2]CHEN S,DOOLEN G D.Lattice Boltzmann Method for Fluid Flows[J].Annual Review of Fluid Mechanics,1998,30(1):329-364.
[3]AN B,SANG W M.The numerical stu-dy of lattice Boltzmann method based on different grid structure[J].Chinese Journal of Theoretical and Applied Mechanics,2013,45(5):699-706.
[4]AN B,MENG X Y,YANG S J,et al.Research on the latticeBoltzmann algorithm for grid refinement based on non-uniform rectangular grid[J].Chinese Journal of Theoretical and Applied Mechanics,2023,55(10):2288-2296
[5]MARVRIPLIS D J.Multigrid solution of the steady state lattice Boltzmann equation[J].Computers & Fluids,2006,35:793-804.
[6]PATIL D V,PREMNATH K N,BANERJEE S.Multigrid lattice Boltzmann method for accelerated solution of elliptic equations[J].Journal of Computational Physics,2014,265:172-194.
[7]FILIPPOVA O,HÄNEL D.Grid Refinement for Lattice-BGK Models[J].Journal of Computational Physics,1998,147(1):219-228.
[8]DUPUIS A,CHOPARD B.Theory and applications of an alternative lattice Boltzmann grid refinement algorithm[J].Physical Review E,2003,67(6):066707.
[9]LAGRAVA D,MALASPINAS O,LATT J,et al.Advances inmulti-domain lattice Boltzmann grid refinement[J].Journal of Computational Physics,2012,231(14):4808-4822.
[10]LATT J,MALASPINAS O,KONTAXAKIS D,et al.Palabos:Parallel Lattice Boltzmann Solver[J].Computers & Mathema-tics with Applications,2021,81:334-350.
[11]HASERT M,MASILAMANI K,ZIMNY S,et al.Complex fluid simulations with the parallel tree-based Lattice Boltzmann solver Musubi[J].Journal of Computational Science,2014,5(5):784-794.
[12]FEICHTINGER C,DONATH S,KÖSTLER H,et al.WaLBerla:HPC software design for computational engineering simulations[J].Journal of Computational Science,2011,2(2):105-112.
[13]ROHDE M,KANDHAI D,DERKSEN J J,et al.A generic,mass conservative local grid refinement technique for lattice-Boltzmann schemes[J].Numerical Methods in Fluids,2006,51(4):439-468.
[14]KARYPIS G,KUMAR V.A fast and high quality multilevelscheme for partitioning irregular graphs[J].SIAM Journal on Scientific Computing,1998,20(1):359-392.
[15]CAMPBELL P M,DEVINE K D,FLAHERTY J E,et al.TERESCO,Dynamic octree load balancing using space-filling curves[R].Williams College Department of Computer Science Technical Report.2003.
[16]MORTON G M.A computer oriented geodetic data base and a new technique in the file sequencing[J]International Business Machines Ltd.,1966,20.
[17]BORRELL R,OYARZUN G,DOSIMONT D,et al.ParallelSFC-based mesh partitioning and load balancing[C]// 2019 IEEE/ACM 10th Workshop on Latest Advances in Scalable Algorithms for Large-Scale Systems(ScalA).2019:72-78.
[18]WHITE A T,CHONG C K.Rotational invariance in the three-dimensional lattice Boltzmann method is dependent on the choice of lattice[J].Journal of Computational Physics,2011,230(16):6367-6378.
[19]GEIER M,SCHÖNHERR M,PASQUALI A,et al.The cumulant lattice Boltzmann equation in three dimensions:Theory and validation[J].Computers & Mathematics with Applications,2015,70(4):507-547.
[20]SIMON H D,TENG S H.How Good is Recursive Bisection?[J].SIAM Journal of Scientific Computing,1997,18(5):1436-1445.
[21]DRIESSCHE R V,ROOSE D.An improved spectral bisectionalgorithm and its application to dynamic load balancing[J].Parallel Computing,1995,21(1):29-48.
[22]CLIFT R,GRACE J R,WEBER M E.Bubbles,drops,and particles[M].Academic Press,1978.
[23]HEFT A I,INDINGER T,ADAMS N A.Introduction of a new realistic generic car model for aerodynamic investigations[R].SAE Technical Paper,2012.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发