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 2209.01476cs.LGcond-mat.dis-nn
收藏

基于拉格朗日图神经网络的粒子系统动力学学习

Learning the Dynamics of Particle-based Systems with Lagrangian Graph Neural Networks

Indian Institute of Technology Delhi(印度理工学院德里分校)

Ravinder Bhattoo, Sayan Ranu, N. M. Anoop Krishnan

摘要:

物理系统通常被表示为粒子的组合,粒子的个体动力学支配系统动力学。然而,传统的方法需要几个抽象的量,如能量或力的知识来推断这些粒子的动力学。在这里,我们提出了一个框架,即拉格朗日图神经网络(LGnn),它提供了一个强大的归纳偏差,直接从轨迹学习基于粒子的系统的拉格朗日。我们在具有约束和阻力的具有挑战性的系统上测试了我们的方法- LGnn优于基线,如前馈拉格朗日神经网络(Lnn),性能有所改善。我们还显示了系统的zero-shot泛化能力,通过模拟比训练的系统大两个数量级的系统,以及模型看不到的混合系统,这是一个独特的功能。与Lnn相比,LGnn的图形架构显著简化了学习,在数据量减少约20倍的情况下,性能提高约25倍。最后,我们展示了LGnn的可解释性,它直接提供了模型学习的阻力和约束力的物理见解。因此,LGnn可以为纯粹从可观测量理解物理系统的动力学提供一个刺激。

英文摘要:

Physical systems are commonly represented as a combination of particles, the individual dynamics of which govern the system dynamics. However, traditional approaches require the knowledge of several abstract quantities such as the energy or force to infer the dynamics of these particles. Here, we present a framework, namely, Lagrangian graph neural network (LGnn), that provides a strong inductive bias to learn the Lagrangian of a particle-based system directly from the trajectory. We test our approach on challenging systems with constraints and drag -- LGnn outperforms baselines such as feed-forward Lagrangian neural network (Lnn) with improved performance. We also show the zero-shot generalizability of the system by simulating systems two orders of magnitude larger than the trained one and also hybrid systems that are unseen by the model, a unique feature. The graph architecture of LGnn significantly simplifies the learning in comparison to Lnn with ~25 times better performance on ~20 times smaller amounts of data. Finally, we show the interpretability of LGnn, which directly provides physical insights on drag and constraint forces learned by the model. LGnn can thus provide a fillip toward understanding the dynamics of physical systems purely from observable quantities.

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。

↑