In this paper, we propose a new adaptive technique, named adaptive trajectories sampling (ATS), which is used to select training points for the numerical solution of partial differential equations (PDEs) with deep learning methods. The key feature of the ATS is that all training points are adaptively selected from trajectories that are generated according to a PDE-related stochastic process. We incorporate the ATS into three known deep learning solvers for PDEs, namely the adaptive derivative-free-loss method (ATS-DFLM), the adaptive physics-informed neural network method (ATS-PINN), and the adaptive temporal-difference method for forward-backward stochastic differential equations (ATS-FBSTD). Our numerical experiments demonstrate that the ATS remarkably improves the computational accuracy and efficiency of the original deep learning solvers for the PDEs. In particular, for some specific high-dimensional PDEs, the ATS can even improve the accuracy of the PINN by two orders of magnitude.
翻译:本文提出一种名为自适应轨迹采样(ATS)的新的自适应技术,用于选择训练点以便使用深度学习方法数值求解偏微分方程(PDE)。ATS的关键特点在于,所有训练点都是从生成的与PDE相关的随机过程的轨迹中自适应选择的。我们将ATS纳入三种已知的PDE深度学习求解器中,分别是自适应无导数损失方法(ATS-DFLM)、自适应基于物理法的神经网络方法(ATS-PINN)以及用于正反向随机微分方程的自适应时序差分方法(ATS-FBSTD)。我们的数值实验表明,ATS显著提高了原始深度学习求解器对于PDE的计算精度和计算效率。特别地,对于某些特定的高维PDE,ATS甚至能够将PINN的精度提高两个数量级。