WebGL技术下的SPH流体模拟方法
|
摘要
随着WebGL技术标准的发展,浏览器端可以无插件地进行复杂实时的计算机流体模拟,但是流体的实时模拟对平台的实时计算和渲染能力有很高的要求,同时常用的流体模拟方法在Web前端很难达到理想的渲染帧率.针对该问题,提出了在WebGL环境下基于多帧缓冲区的SPH流体实时模拟方法,首先,对传统的邻域粒子链表搜索算法进行改进;然后在频繁的粒子位置与速度更新过程中,提出基于多帧缓冲区的中间数据存储和传输方法;最后,采用离屏渲染的方式将多个需要绘制的数据一次性送入主帧缓冲区中进行渲染.实验结果表明,该方法可以使SPH流体模拟完全运行在GPU上,并且该方法是基于CPU的链表搜索算法的渲染帧率的5倍,同时,在粒子数目增加至50k且进一步增加的情况下,比基于GPU的Bayraktar算法具有明显的优势,能够满足在Web前端进行三维场景渲染的帧率要求.
With the development of WebGL technical standards,we can realize the complex and real-time fluid simulation rendering with no plugins in Web front-end.But the real-time calculation and rendering capabilities have high demands.To solve the problem mentioned before,this paper presents a SPH fluid real-time simulation method based on multi-frame buffers under WebGL.Firstly,we improved the traditional particles chain table search algorithm,and proposed to build the particles position number look-up table in grids for redundant computation to determine the neighbor grids of current particle during the algorithm process.Then,we proposed intermediate data storage and transmission method based on multi-frame buffer in the frequent update process of particle position and velocity,which can decrease the data transfer between CPU and GPU.Finally,we put many intermediate results into off-screen frame buffer to blend and used the off-screen rendering method,which can draw a plurality of data at once into the main frame buffer rendering.This method allows SPH fluid simulation runs entirely on the GPU.The experimental results show that rendering frame rate of this method is five times than linked-list search algorithm based on CPU.when the number of particles increases up to 50 K and further,this method has obvious advantages comparing with Bayraktar GPU-based algorithm,and can meet the frame rate requirements of three-dimensional scene rendering in Web front-end.
With the development of WebGL technical standards,we can realize the complex and real-time fluid simulation rendering with no plugins in Web front-end.But the real-time calculation and rendering capabilities have high demands.To solve the problem mentioned before,this paper presents a SPH fluid real-time simulation method based on multi-frame buffers under WebGL.Firstly,we improved the traditional particles chain table search algorithm,and proposed to build the particles position number look-up table in grids for redundant computation to determine the neighbor grids of current particle during the algorithm process.Then,we proposed intermediate data storage and transmission method based on multi-frame buffer in the frequent update process of particle position and velocity,which can decrease the data transfer between CPU and GPU.Finally,we put many intermediate results into off-screen frame buffer to blend and used the off-screen rendering method,which can draw a plurality of data at once into the main frame buffer rendering.This method allows SPH fluid simulation runs entirely on the GPU.The experimental results show that rendering frame rate of this method is five times than linked-list search algorithm based on CPU.when the number of particles increases up to 50 K and further,this method has obvious advantages comparing with Bayraktar GPU-based algorithm,and can meet the frame rate requirements of three-dimensional scene rendering in Web front-end.
引文
[1]Micky Kelager.Lagrangian fluid dynamics using smoothed particle hydrodynamics[D].Copenhagen:University of Copenhagen,2006.
[2]Harada T,Koshizuka S,Kawaguchi Y.Smoothed particle hydrodynamics on GPUS[J].Structure,2007,4(4):671-691.
[3]Lei Feng,Wang Chao-liang,Li Chuang-rong,et al.A research for3D Web GIS based on Web GL[J].International Conference on Computer Science&Network Technology(ICCSNT),2011,1:348-351.
[4]Khronos releases final Web GL 1.0 specification[EB/OL].https://w w w.khronos.org/new s/press/khronos-releases-final-webgl-1.0-specification,2011.
[5]Sloup P.Web GL earth[D].Brno:Masaryk University Faculty of Informatics,2011.
[6]DG Iglesias.Design and implementation of 3D buildings integration for a Web GL-Based virtual globe:a case study of valencian cadastre and fide building mode[D].M asters Program in Geospatial Technologies,Jaime I University,2012.
[7]Hering N,Runz M,Samecki L,et al.3DGIS:a real-time browserrendered 3D campus information system based on Web GL[C].World Comp,2012.
[8]Rez,Jose,Trujillo,et al.An open source virtual globe framework for i OS,Android and Web GL compliant brow ser[C].Association for Computing M achinery(ACM)International Conference Proceeding Series,Association for Computing M achinery,2012:1-10.
[9]Chen Ying-jian.Real-time simulation of fluid in pipes based on SPH methods and the optimization on GPU[D].Qinhuangdao:Yanshan University,2012.
[10]Ding Liang.SPH formula solution of the navier-strokes eauation[J].Information Technology,2013,2(2):170-174.
[11]Yang Xu-bo.Interactive fluid simulation on the GPU[D].Shanghai:Shanghai Jiaotong University,2014.
[12]He Jian.Study on the real-time fluid simulation with complex boundaries[D].Hangzhou:Zhejiang University,2010.
[13]Matthias Muller,David Charypar,Markus Gross.Particle-based fluid simulation for interactive application[C].Association for Computing M achinery(ACM)SIGGRAPH/Eurographics Symposium on Computer Animation,Association for Computing M achinery,2003:154-159.
[14]Lin Hao.Open GL visualization based on the smoothed particle hydrodynamics method[D].Hefei:Anhui University,2011.
[15]Zhao Xiang-kun,Li Feng-xia,Zhan Shou-yi.GPU-based neighbor search algorithm for SPH fluid simulations[J].Journal of South China University of Technology(Natural Science Edition),2011,39(7):150-155.
[16]Swrkan Bayraktar,Ugur Gudukbay,Bulent Ozguc.GPU-based neighbor-search algorithm for particle simulations[J].Journal of Graphics GPU&Game Tools,2009,14(1):31-42.
[17]Tang Yong,Chen Ying-jian,Lv Meng-ya.A multidimensional tree searching algorithm by SPH method on GPU[J].Journal of Chinese Computer Systems,2013,34(12):2833-2836.
[18]Guo Qiu-lei,Tang Yi-zhi,Liu Shi-qiu,et al.A fluid GPU implementation framework of SPH fluid real-time simulation[J].Computer Application and Software,2011,28(11):69-72.
[19]Zhou Xiao-nan,Zhou Gui-yun.SPH fluid flow simulation[J].Geospatial Information,2015,13(2):86-90.
[20]Xia Xi-lin,Liang Qiu-hua.A GPU-accelerated smoothed particle hydrodynamics(SPH)model for the shallow water equations[J].Environmental M odelling&Software,2016,75:28-43.
[21]SPH algorithm Introduction:smooth kernel function[EB/OL].Http://www.starming.com/index.php?action=plugin&v=wave&tpl=union&ac=viewgrouppost&gid=34694&tid=12522,2011.
[22]Zhou Li-bo.Research on visualization of 3D virtual earth based on Web GL[D].Beijing:China University of Geosciences,2013.
[23]Wen Chan-juan,Ou Jia-wei,Jia Jin-yuan.GPGPU-based smoothed particle hydrodynamic fluid simulation[J].Journal of ComputerAided Design&Computer Graphics,2010,22(3):406-411.
[24]Pharr Matt,Fernando Randima.GPU gems 2:Programming techniques for high-perfonmance graphics and general purpose computation[M].Boston:Addison-Wesley,2005.
[25]Fenando Randima.GPU gems:programming techniques,tips and tricks for real-time graphics[M].Boston:Addison-Wesley,2005.
[26]Cedeman D,Tsigas P.A practical quick sort algorighm for graphics processors[J].Algorithms-ESA 2008,Lecture Notes in Computer Science,Vol 5193.Springer Berlin Heidelberg,2008,14:246-258.
[27]Dong Z,Chen W,Bao H,et al.Real-time voxelization for complex models[C].Computer Graphics&Applications,Pacific Conference,PC,2015:43-50.
[28]Stefano Sibilla.An algorithm to improve consisitency in smoothed particle hydrodynamics[J].Computer&Fluids,2015,118:148-158.
[29]William E Lorensen,Harvey E Cline.Marching cubes:a high resolution 3D surface construction algorithm[J].Acm Siggraph Computer Graphics,1987,21(4):163-169.
[30]ystein E.Krog,Anne C Elster.Fast GPU-based fluid simulations using SPH[J].Applied Parallel&Scientific Computing.Springer Berlin Heidelberg,2010:98-109.
[9]陈英建.SPH实现管道流体实时可视化及在GPU上的优化[D].秦皇岛:燕山大学,2012.
[10]丁亮.Navier-Stokes方程SPH的公式解法[J].信息技术,2013,2(2):170-174.
[11]杨旭波.基于GPU的可交互实时流体模拟[D].上海:上海交通大学,2014.
[12]何戬.复杂流体场景的实时模拟研究[D].杭州:浙江大学,2010.
[14]林昊.基于光滑粒子流体动力学法的Open GL可视化[D].合肥:安徽大学,2011.
[15]赵相坤,李凤霞,战守义.基于GPU的面向SPH流体模拟的邻居查找算法[J].华南理工大学学报(自然科学版),2011,39(7):150-155.
[17]唐勇,陈英建,吕梦雅.SPH方法中多维树搜索算法的优化及在GPU上的实现[J].小型微型计算机系统,2013,34(12):2833-2836.
[19]周枭楠,周贵云.基于DEM数据的SPH水体流动模拟[J].地理空间信息,2015,13(2):86-90.
[22]周立博.基于Web GL的三维虚拟地球可视化研究[D].北京:中国地质大学,2013.
[23]温婵娟,欧嘉蔚,贾金原.GPU通用计算平台上的SPH流体模拟[J].计算机辅助设计与图形学学报,2010,22(3):406-411.
[2]Harada T,Koshizuka S,Kawaguchi Y.Smoothed particle hydrodynamics on GPUS[J].Structure,2007,4(4):671-691.
[3]Lei Feng,Wang Chao-liang,Li Chuang-rong,et al.A research for3D Web GIS based on Web GL[J].International Conference on Computer Science&Network Technology(ICCSNT),2011,1:348-351.
[4]Khronos releases final Web GL 1.0 specification[EB/OL].https://w w w.khronos.org/new s/press/khronos-releases-final-webgl-1.0-specification,2011.
[5]Sloup P.Web GL earth[D].Brno:Masaryk University Faculty of Informatics,2011.
[6]DG Iglesias.Design and implementation of 3D buildings integration for a Web GL-Based virtual globe:a case study of valencian cadastre and fide building mode[D].M asters Program in Geospatial Technologies,Jaime I University,2012.
[7]Hering N,Runz M,Samecki L,et al.3DGIS:a real-time browserrendered 3D campus information system based on Web GL[C].World Comp,2012.
[8]Rez,Jose,Trujillo,et al.An open source virtual globe framework for i OS,Android and Web GL compliant brow ser[C].Association for Computing M achinery(ACM)International Conference Proceeding Series,Association for Computing M achinery,2012:1-10.
[9]Chen Ying-jian.Real-time simulation of fluid in pipes based on SPH methods and the optimization on GPU[D].Qinhuangdao:Yanshan University,2012.
[10]Ding Liang.SPH formula solution of the navier-strokes eauation[J].Information Technology,2013,2(2):170-174.
[11]Yang Xu-bo.Interactive fluid simulation on the GPU[D].Shanghai:Shanghai Jiaotong University,2014.
[12]He Jian.Study on the real-time fluid simulation with complex boundaries[D].Hangzhou:Zhejiang University,2010.
[13]Matthias Muller,David Charypar,Markus Gross.Particle-based fluid simulation for interactive application[C].Association for Computing M achinery(ACM)SIGGRAPH/Eurographics Symposium on Computer Animation,Association for Computing M achinery,2003:154-159.
[14]Lin Hao.Open GL visualization based on the smoothed particle hydrodynamics method[D].Hefei:Anhui University,2011.
[15]Zhao Xiang-kun,Li Feng-xia,Zhan Shou-yi.GPU-based neighbor search algorithm for SPH fluid simulations[J].Journal of South China University of Technology(Natural Science Edition),2011,39(7):150-155.
[16]Swrkan Bayraktar,Ugur Gudukbay,Bulent Ozguc.GPU-based neighbor-search algorithm for particle simulations[J].Journal of Graphics GPU&Game Tools,2009,14(1):31-42.
[17]Tang Yong,Chen Ying-jian,Lv Meng-ya.A multidimensional tree searching algorithm by SPH method on GPU[J].Journal of Chinese Computer Systems,2013,34(12):2833-2836.
[18]Guo Qiu-lei,Tang Yi-zhi,Liu Shi-qiu,et al.A fluid GPU implementation framework of SPH fluid real-time simulation[J].Computer Application and Software,2011,28(11):69-72.
[19]Zhou Xiao-nan,Zhou Gui-yun.SPH fluid flow simulation[J].Geospatial Information,2015,13(2):86-90.
[20]Xia Xi-lin,Liang Qiu-hua.A GPU-accelerated smoothed particle hydrodynamics(SPH)model for the shallow water equations[J].Environmental M odelling&Software,2016,75:28-43.
[21]SPH algorithm Introduction:smooth kernel function[EB/OL].Http://www.starming.com/index.php?action=plugin&v=wave&tpl=union&ac=viewgrouppost&gid=34694&tid=12522,2011.
[22]Zhou Li-bo.Research on visualization of 3D virtual earth based on Web GL[D].Beijing:China University of Geosciences,2013.
[23]Wen Chan-juan,Ou Jia-wei,Jia Jin-yuan.GPGPU-based smoothed particle hydrodynamic fluid simulation[J].Journal of ComputerAided Design&Computer Graphics,2010,22(3):406-411.
[24]Pharr Matt,Fernando Randima.GPU gems 2:Programming techniques for high-perfonmance graphics and general purpose computation[M].Boston:Addison-Wesley,2005.
[25]Fenando Randima.GPU gems:programming techniques,tips and tricks for real-time graphics[M].Boston:Addison-Wesley,2005.
[26]Cedeman D,Tsigas P.A practical quick sort algorighm for graphics processors[J].Algorithms-ESA 2008,Lecture Notes in Computer Science,Vol 5193.Springer Berlin Heidelberg,2008,14:246-258.
[27]Dong Z,Chen W,Bao H,et al.Real-time voxelization for complex models[C].Computer Graphics&Applications,Pacific Conference,PC,2015:43-50.
[28]Stefano Sibilla.An algorithm to improve consisitency in smoothed particle hydrodynamics[J].Computer&Fluids,2015,118:148-158.
[29]William E Lorensen,Harvey E Cline.Marching cubes:a high resolution 3D surface construction algorithm[J].Acm Siggraph Computer Graphics,1987,21(4):163-169.
[30]ystein E.Krog,Anne C Elster.Fast GPU-based fluid simulations using SPH[J].Applied Parallel&Scientific Computing.Springer Berlin Heidelberg,2010:98-109.
[9]陈英建.SPH实现管道流体实时可视化及在GPU上的优化[D].秦皇岛:燕山大学,2012.
[10]丁亮.Navier-Stokes方程SPH的公式解法[J].信息技术,2013,2(2):170-174.
[11]杨旭波.基于GPU的可交互实时流体模拟[D].上海:上海交通大学,2014.
[12]何戬.复杂流体场景的实时模拟研究[D].杭州:浙江大学,2010.
[14]林昊.基于光滑粒子流体动力学法的Open GL可视化[D].合肥:安徽大学,2011.
[15]赵相坤,李凤霞,战守义.基于GPU的面向SPH流体模拟的邻居查找算法[J].华南理工大学学报(自然科学版),2011,39(7):150-155.
[17]唐勇,陈英建,吕梦雅.SPH方法中多维树搜索算法的优化及在GPU上的实现[J].小型微型计算机系统,2013,34(12):2833-2836.
[19]周枭楠,周贵云.基于DEM数据的SPH水体流动模拟[J].地理空间信息,2015,13(2):86-90.
[22]周立博.基于Web GL的三维虚拟地球可视化研究[D].北京:中国地质大学,2013.
[23]温婵娟,欧嘉蔚,贾金原.GPU通用计算平台上的SPH流体模拟[J].计算机辅助设计与图形学学报,2010,22(3):406-411.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |