Discovery of Physics From Data: Universal Laws and Discrepancies.
Discovery of Physics From Data: Universal Laws and Discrepancies.
复制标题
从数据中发现物理学：普遍法律和差异。
DOI：
10.3389/frai.2020.00025
复制
发表时间：
2020
影响因子：
4
通讯作者：
Kutz JN
中科院分区：
文献类型：
--
作者：
de Silva BM;Higdon DM;Brunton SL;Kutz JN
机器学习（ML）和人工智能（AI）算法现在被用于从测量数据中自动发现物理原理和控制方程。然而，如果不同时提出一个伴随的差异模型来解释理论和测量之间不可避免的不匹配，那么从数据中假设一个普遍的物理定律是具有挑战性的。通过重新审视不同大小和质量的落体建模的经典问题，我们强调了一些微妙的问题，必须通过现代数据驱动的自动化物理发现方法来解决。具体来说，我们表明测量噪声和复杂的次要物理机制，如非定常流体阻力，可以模糊潜在的万有引力定律，导致错误的模型。我们使用非线性动力学的稀疏识别（SINDy）方法来识别实际测量数据和模拟轨迹的控制方程。在SINDy中加入一个假设，即每个落体都受相似的物理定律支配，这被证明可以提高学习模型的鲁棒性，但由于阻力动力学的微妙性，预测和观察之间的差异仍然存在。这项工作强调了这样一个事实，即未经进一步修改，ML/AI的幼稚应用通常不足以推断普遍的物理定律。
Machine learning (ML) and artificial intelligence (AI) algorithms are now being used to automate the discovery of physics principles and governing equations from measurement data alone. However, positing a universal physical law from data is challenging without simultaneously proposing an accompanying discrepancy model to account for the inevitable mismatch between theory and measurements. By revisiting the classic problem of modeling falling objects of different size and mass, we highlight a number of nuanced issues that must be addressed by modern data-driven methods for automated physics discovery. Specifically, we show that measurement noise and complex secondary physical mechanisms, like unsteady fluid drag forces, can obscure the underlying law of gravitation, leading to an erroneous model. We use the sparse identification of non-linear dynamics (SINDy) method to identify governing equations for real-world measurement data and simulated trajectories. Incorporating into SINDy the assumption that each falling object is governed by a similar physical law is shown to improve the robustness of the learned models, but discrepancies between the predictions and observations persist due to subtleties in drag dynamics. This work highlights the fact that the naive application of ML/AI will generally be insufficient to infer universal physical laws without further modification.
登录
查看更多内容
影响因子：
0.9
作者：
ADLER, CG;COULTER, BL
通讯作者：
COULTER, BL
影响因子：
1
作者：
Falconer, Isobel
通讯作者：
Falconer, Isobel
影响因子：
1.2
作者：
MAGARVEY, RH;MACLATCHY, CS
通讯作者：
MACLATCHY, CS
影响因子：
3.7
作者：
ACHENBACH, E
通讯作者：
ACHENBACH, E
影响因子：
5.7
作者：
Breiman, L
通讯作者：
Breiman, L