首页> 外文会议>ACS National Meeting Exhibition >NEURAL NETWORK POTENTIALS FOR LARGE-SCALE MOLECULAR DYNAMICS SIMULATIONS OF CONDENSED SYSTEMS
【24h】

NEURAL NETWORK POTENTIALS FOR LARGE-SCALE MOLECULAR DYNAMICS SIMULATIONS OF CONDENSED SYSTEMS

机译:凝聚系统大规模分子动力学模拟的神经网络电位

获取原文

摘要

Ab initio molecular dynamics (MD) simulations allow studying a wide range of systems with good accuracy: Unfortunately, as the electronic structure problem is solved 'on-the-fly' in.each time step, typically employing density-functional theory (DFT), these simulations are computationally too demanding to address a variety of interesting problems. Consequently, more efficient potentials are heeded to perform extended MD simulations of large systems. To solve this problem, many types of potentials have been suggested in the literature in the past decades, from more approximate solutions of the electronic structure problem to classical force fields. While these potentials allow studying larger systems, the accuracy is necessarily limited by the underlying approximations.
机译:AB Initio分子动力学(MD)模拟允许研究具有良好精度的各种系统:遗憾的是,随着电子结构问题在“在飞行”中解决了“在飞行”中,通常采用密度 - 功能理论(DFT) ,这些模拟是计算的太大要求,以解决各种有趣的问题。因此,HEEDED的更有效的潜力以执行大型系统的扩展MD仿真。为了解决这个问题,在过去几十年中,在文献中提出了许多类型的潜力,从电子结构问题对古典力领域的更近似解。虽然这些潜力允许研究较大的系统,但精度必须受到底层近似的限制。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号