Login / Signup

Unified Efficient Thermostat Scheme for the Canonical Ensemble with Holonomic or Isokinetic Constraints via Molecular Dynamics.

Zhijun ZhangXinzijian LiuKangyu YanMark E TuckermanJian Liu
Published in: The journal of physical chemistry. A (2019)
We have recently proposed a new unified theoretical scheme (the "middle" scheme) for thermostat algorithms for efficient and accurate configurational sampling of the canonical ensemble. In this paper, we extend the "middle" scheme to molecular dynamics algorithms for configurational sampling in systems subject to constraints. Holonomic constraints and isokinetic constraints are used for demonstration. Numerical examples indicate that the "middle" scheme presents a promising approach to calculate configuration-dependent thermodynamic properties and their thermal fluctuations.
Keyphrases
  • molecular dynamics
  • density functional theory
  • machine learning
  • visible light
  • deep learning
  • convolutional neural network
  • neural network