arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2502.05201cond-mat.mtrl-scicond-mat.dis-nn
收藏

利用人工智能对非晶态及非晶态多孔钯进行建模:一种深度神经网络方法

Harnessing Artificial Intelligence for Modeling Amorphous and Amorphous Porous Palladium: A Deep Neural Network Approach

Isaías Rodríguez

AI总结:

本文提出用深度神经网络拟合从头算分子动力学数据来建模非晶态及非晶态多孔钯的原子间势,在保证精度接近密度泛函理论的同时大幅降低计算成本,为加速非晶材料在能源与催化中的应用提供了新途径。

摘要:

非晶态及非晶态多孔钯是催化、储氢和功能应用中的关键材料,但其复杂结构带来了计算挑战。本研究采用一个深度神经网络,该网络基于从头算分子动力学模拟得到的33,310个原子构型进行训练,以对其原子间势进行建模。这种人工智能驱动的方法能够准确预测结构和热学性质,同时显著降低计算成本。与密度泛函理论的对比验证证实了其在重现力、能量和结构分布方面的可靠性。这些发现凸显了人工智能在加速非晶态材料研究以及推动其在能源和催化领域应用方面的潜力。

英文摘要:

Amorphous and amorphous porous palladium are key materials for catalysis, hydrogen storage, and functional applications, but their complex structures present computational challenges. This study employs a deep neural network trained on 33,310 atomic configurations from ab initio molecular dynamics simulations to model their interatomic potential. The AI-driven approach accurately predicts structural and thermal properties while significantly reducing computational costs. Validation against density functional theory confirms its reliability in reproducing forces, energies, and structural distributions. These findings highlight AI's potential in accelerating the study of amorphous materials and advancing their applications in energy and catalysis.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑