物理嵌入和机器学习的协同集成,使精确可靠的力场能够实现
Lifeng Xu1,2, Jian Jiang1,2
1Beijing National Laboratory for Molecular Sciences, State Key Laboratory of Polymer Physics and Chemistry, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, P. R. China.
这项研究引入了一个物理信息的神经网络 (PINN) 力量场,将物理原理与机器学习相结合,用于准确的分子模拟. 这种新的方法可以在最小的计算成本下实现高精度和宏观性质的强大预测.
科学领域:
- 计算化学是一种计算化学.
- 机器学习在材料科学中的应用
背景情况:
- 机器学习力场 (MLFFs) 提供了量子化学准确性,但与外推和远程相互作用扎.
- 现有的MLFF在预测宏观性质和导航新化学空间方面面临挑战.
研究的目的:
- 通过协同整合物理原理和机器学习来开发一个物理信息的神经网络 (PINN) 力量场.
- 为了解决当前MLFFs在化学空间外推,静电相互作用和宏观性质预测方面的局限性.
主要方法:
- 在PINN框架内将物理知识纳入神经网络参数.
- 采用了新的塔布-亚当算法,在物理约束下进行高效的全球优化.
- 使用AMOEBA+力场作为基于物理的模型,并在二乙烯基醇二甲基乙醇 (DEGDME) 数据集上进行训练/测试.
主要成果:
- 实现了精确且耐噪声的机器学习力场.
- 在描述分子相互作用方面表现出了显著的概括和密度函数理论 (DFT) 准确性.
- 能够准确预测宏观性质,如扩散系数,并降低计算成本.
结论:
- 开发的PINN力场代表了在构建准确和强大的MLFFs方面取得的突破.
- 这种方法为未来开发化学中物理信息化的机器学习模型提供了基本框架.
- 该研究强调了将物理原理与机器学习结合起来,实现高效准确的分子模拟的潜力.
更多相关视频
07:31Author Spotlight: Advancing Cell Membrane Biophysics - Exploring Interactions and Challenges Through Experimental and Computational Approaches
Published on: September 1, 2023
06:45Force Spectroscopy of Single Protein Molecules Using an Atomic Force Microscope
Published on: February 28, 2019
相关概念视频
Three-Dimensional Force System
Three-Dimensional Force System:Problem Solving
To solve a three-dimensional force system, first resolve each force into its respective scalar components. Do this using...
Two-Dimensional Force System
Two-Dimensional Force System: Problem Solving
The first step to solving a two-dimensional force system problem is to draw a free-body diagram of the object under consideration. This diagram helps identify all the external forces acting on the object, including their...
Static and Kinetic Frictional Force
However, if two systems are in contact and are stationary relative to one...
Atomic Force Microscopy
The AFM Probe
The probe is regarded as the heart of any AFM setup and comprises the...
