Application to a New AMG Method
- 463 Downloads
This chapter presents a new AMG method for nonsymmetric problems involving heuristics inspired directly by the convergence theory of the previous chapter. While less rigorous than previous chapters, it proves by example that the abstract theory of previous chapters can be used to inform, quite directly, the development of practical methods. The presented heuristics include a novel coarsening strategy and interpolation targeting the “approximation properties” of the convergence theory that generalizes the “energy-minimizing” weights used in symmetric methods. Numerical results are presented, showing how well the developed method works, and also demonstrating how well the bounds of the theory work on a real example.
KeywordsPrevious Chapter Column Norm GMRES Iteration Jacobi Iteration Convergence Bound
- 3.Brannick, J., Zikatanov, L.: Algebraic multigrid methods based on compatible relaxation and energy minimization. In: Proceedings of the 16th International Conference on Domain Decomposition Methods (2005)Google Scholar
- 9.Nikiforov, V.: Revisiting schur’s bound on the largest singular value (2007). http://arXiv.org/abs/math/0702722