Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems
Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems
Jeremy Yu,Lu Lu,Xuhui Meng,G. Karniadakis
2021 · DOI: 10.1016/j.cma.2022.114823
Computer Methods in Applied Mechanics and Engineering · 引用 572 次
TLDR
This work proposes a new method, gradient-enhanced physics-informed neural networks (gPINNs), for improving the accuracy and training efficiency of PINNs, and combines the method of residual-based adaptive refinement (RAR) to further improve the performance of gPINN, especially in PDEs with solutions that have steep gradients.