Skip to content

PINN solution fits data points but diverges from exact solution for 1D wave equation #1

Description

@Patience-ojoboh

Issue Description:
Description: I am working on implementing a Physics-Informed Neural Network (PINN) in PyTorch to solve the 1D acoustic wave equation. My goal is to train the PINN to approximate the solution of the wave equation using a combination of boundary conditions and physics loss (PDE residual). However, despite my efforts, the PINN prediction tends to align with the physics and boundary data points but diverges from the exact solution.
PINN.zip

Issue: While training, the PINN prediction closely fits the boundary and physics points, but it does not generalize well to match the exact solution of the 1D wave equation across the entire domain. The issue persists despite increasing the number of training iterations and adjusting hyperparameters such as the weight of the physics loss term.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions