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 2502.04495cs.LG
收藏

通过不变函数学习发现动力系统的物理定律

Discovering Physics Laws of Dynamical Systems via Invariant Function Learning

Department of Computer Science \& Engineering, Texas A\&M University, College Station, Texas, USA

Shurui Gui, Xiner Li, Shuiwang Ji

AI总结:

针对复杂环境下跨环境发现ODE动力系统内在规律的难题,提出基于因果分析的DIF方法,通过编码器-解码器超网络解耦不变函数与环境特定动力学,实验验证其有效性与高效性,可揭示系统内在物理定律。

摘要:

我们研究学习由常微分方程(ODE)支配的动力系统的内在规律。核心挑战在于如何在规避环境特定机制的同时,跨多个环境发现内在动力学。与现有工作不同,我们处理更复杂的环境:其中的变化不仅限于函数系数,还延伸至完全不同的函数形式。例如,我们通过观测不同环境下的摆动力学(如阻尼环境中的$α^2 \sin(θ_t) - ρω_t$和驱动环境中的$α^2 \sin(θ_t) + ρ\frac{ω_t}{\left|ω_t\right|}$),证明了理想摆自然运动$α^2 \sin{θ_t}$的发现。我们将该问题表述为不变函数学习任务,并提出一种基于因果分析的新方法——不变函数解耦(Disentanglement of Invariant Functions, DIF)。我们构建了因果图,并设计了一个编码器-解码器超网络,可显式将不变函数与环境特定动力学解耦。我们提出的基于信息的原理强制提取的不变函数与环境相互独立,为不变函数的发现提供了保障。在三个ODE系统上与元学习、不变学习基线的定量比较,证明了我们方法的有效性和高效性。此外,符号回归解释结果凸显了我们的框架揭示内在规律的能力。我们的代码已作为AIRS库的一部分发布,地址为https://github.com/divelab/AIRS/tree/main/OpenODE/DIF。

英文摘要:

We consider learning underlying laws of dynamical systems governed by ordinary differential equations (ODE). A key challenge is how to discover intrinsic dynamics across multiple environments while circumventing environment-specific mechanisms. Unlike prior work, we tackle more complex environments where changes extend beyond function coefficients to entirely different function forms. For example, we demonstrate the discovery of ideal pendulum's natural motion $α^2 \sin{θ_t}$ by observing pendulum dynamics in different environments, such as the damped environment $α^2 \sin(θ_t) - ρω_t$ and powered environment $α^2 \sin(θ_t) + ρ\frac{ω_t}{\left|ω_t\right|}$. Here, we formulate this problem as an \emph{invariant function learning} task and propose a new method, known as \textbf{D}isentanglement of \textbf{I}nvariant \textbf{F}unctions (DIF), that is grounded in causal analysis. We propose a causal graph and design an encoder-decoder hypernetwork that explicitly disentangles invariant functions from environment-specific dynamics. The discovery of invariant functions is guaranteed by our information-based principle that enforces the independence between extracted invariant functions and environments. Quantitative comparisons with meta-learning and invariant learning baselines on three ODE systems demonstrate the effectiveness and efficiency of our method. Furthermore, symbolic regression explanation results highlight the ability of our framework to uncover intrinsic laws. Our code has been released as part of the AIRS library (\href{https://github.com/divelab/AIRS/tree/main/OpenODE/DIF}{https://github.com/divelab/AIRS/}).

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。

↑