物理化学学报 >> 2009, Vol. 25 >> Issue (03): 430-434.doi: 10.3866/PKU.WHXB20090306

研究论文 上一篇    下一篇

流体系统模拟中邻区列表算法的优化理论

侯吉旋 司黎明   

  1. Laboratoire de Physique, UMR 5182 CNRS, Ecole Normale Superieure de Lyon, 46, Allee d'Italie F-69364 Lyon Cedex 07, France; 北京理工大学信息科学与技术学院电子工程系, 北京 100081
  • 收稿日期:2008-08-25 修回日期:2008-11-04 发布日期:2009-03-02
  • 通讯作者: 司黎明 E-mail:siliming100@yahoo.com.cn

Optimization Theory for Neighbor List Algorithmin Fluid System Simulation

 HOU Ji-Xuan, SI Li-Ming   

  1. Laboratoire de Physique, UMR 5182 CNRS, Ecole Normale Superieure de Lyon, 46, Allee d'Italie F-69364 Lyon Cedex 07, France; Department of Electronic Engineering, School of Information Science and Technology, Beijing Institute of Technology, Beijing 100081, P. R. China
  • Received:2008-08-25 Revised:2008-11-04 Published:2009-03-02
  • Contact: SI Li-Ming E-mail:siliming100@yahoo.com.cn

摘要:

分子动力学模拟中邻区列表算法的效率依赖于其参数的选择. 作者提供了一种选择最优化参数的计算方法, 通过分别使用自由粒子近似和扩散近似对所需模拟的时间进行计算, 再对两种近似计算进行比较. 结果表明, 在密度较低或者皮肤半径较小的情况下需要采用自由粒子近似, 而当密度较高或者皮肤半径较大的情况下则需要采用扩散近似. 该方法的结果与Lennard-Jones流体系统的模拟结果符合得很好.

关键词: 计算技术, 分子动力学计算, 流体系统

Abstract:

The efficiency of the neighbor list algorithm in molecular dynamics simulation depends on the parameters chosen. By using the free-particle approximation and the diffusion approximation we can calculate the central processing unit (CPU) time that is used for the simulation. The free-particle approximation can be used in the case of low density or a small skin radius while the diffusion approximation can be used in the case of high density or a large skin radius. Combining the results of these two approximations optimal parameters may be selected and thus CPU time can be saved. Our result coincides with the result of the simulation based on Lennard-Jones fluid systems.

Key words: Computational techniques, Molecular dynamics calculation, Fluid system

MSC2000: 

  • O641