Initialization-enhanced Physics-Informed Neural Network with Domain Decomposition (IDPINN)
This work addresses accuracy issues in PINNs for computational physics, but it appears incremental as it builds on existing PINN methods with specific enhancements.
The authors tackled the problem of improving prediction accuracy in physics-informed neural networks by proposing IDPINN, which enhances initialization and uses domain decomposition, resulting in demonstrated accuracy benefits on forward problems.
We propose a new physics-informed neural network framework, IDPINN, based on the enhancement of initialization and domain decomposition to improve prediction accuracy. We train a PINN using a small dataset to obtain an initial network structure, including the weighted matrix and bias, which initializes the PINN for each subdomain. Moreover, we leverage the smoothness condition on the interface to enhance the prediction performance. We numerically evaluated it on several forward problems and demonstrated the benefits of IDPINN in terms of accuracy.