Corpus ID: 219531529

Accurately Solving Physical Systems with Graph Learning

@article{Shao2020AccuratelySP,
  title={Accurately Solving Physical Systems with Graph Learning},
  author={Han Shao and Tassilo Kugelstadt and Wojciech Palubicki and Jan Bender and Soren Pirk and Dominik Ludewig Michels},
  journal={ArXiv},
  year={2020},
  volume={abs/2006.03897}
}
  • Han Shao, Tassilo Kugelstadt, +3 authors Dominik Ludewig Michels
  • Published 2020
  • Computer Science, Physics, Mathematics
  • ArXiv
  • Iterative solvers are widely used to accurately simulate physical systems. These solvers require initial guesses to generate a sequence of improving approximate solutions. In this contribution, we introduce a novel method to accelerate iterative solvers for physical systems with graph networks (GNs) by predicting the initial guesses to reduce the number of iterations. Unlike existing methods that aim to learn physical systems in an end-to-end manner, our approach guarantees long-term stability… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 60 REFERENCES

    Learning to Simulate Complex Physics with Graph Networks

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Relational inductive biases, deep learning, and graph networks

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Position Based Dynamics

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    Position and orientation based Cosserat rods

    VIEW 4 EXCERPTS

    Position based dynamics

    VIEW 4 EXCERPTS
    HIGHLY INFLUENTIAL

    A Novel CNN-Based Poisson Solver for Fluid Simulation

    VIEW 2 EXCERPTS