高分子塌缩相变和临界吸附相变的计算机模拟和机器学习

Computer simulation and machine learning of polymer collapse and critical adsorption phase transitions

高分子的塌缩和临界吸附是高分子科学中的两个重要相变现象,两者均伴随着高分子构象的显著变化.本文利用朗之万动力学方法和动力学Monte Carlo方法分别模拟了高分子的塌缩和临界吸附,同时获得了不同温度下大量的高分子构象数据.机器学习方法利用模拟得到的大量伸展无规线团态和塌缩液滴态、脱附态和吸附态构象数据训练神经网络,学习高分子不同状态的特征,快速准确地分析不同温度的高分子构象信息,得到对应的塌缩相变温度和临界吸附温度.结果表明机器学习能正确给出高分子体系的相变温度,这为机器学习技术研究高分子的相变提供了新的思路和方法.

Collapse and critical adsorption of polymers are two crucial phase transitions in polymer science,both are accompanied by significant changes in polymer conformation.In this paper,Langevin dynamics and dynamic Monte Carlo methods are used to simulate the collapse and critical adsorption of polymer,respectively,and corresponding phase transition temperatures are estimated.Meanwhile,a large number of polymer conformations at different temperatures are obtained.In the machine learning method,a large number of extended random coil and collapsed spherical,desorption and adsorption conformations are used to train the neural network,so that the neural network can learn the characteristics of different states of the polymer,and it can quickly and accurately analyze the polymer conformations at different temperatures and obtain the corresponding collapse phase transition temperature and critical adsorption temperature.The results demonstrate that machine learning can correctly calculate the phase transition temperature of polymer system,which provides new ideas and methods for machine learning technology in the study of polymer phase transitions.