Chebyshev acceleration of Picard-Lindelöf iteration
This paper complements recent work by \textit{R.D. Skeel} [SIAM J. Sci. Stat. Comput. 10, No. 4, 756-776 (1989; Zbl 0687.65076) and \textit{O. Nevanlinna} [Numer. Math. 57, No. 2, 147-156 (1990; Zbl 0697.65058)] regarding the question as to whether a significant acceleration of waveform iteration (Picard-Lindelöf iteration) is possible. The author shows that suitably defined Chebyshev acceleration of Picard iteration for \(\dot y+Ay=f(t)\), \(0<t<T \leq \infty\), \(y(0)=y_ 0\), yields at least the same error reduction as the usual Chebyshev acceleration of Richardson's iteration for the static linear equation \(x/T+Ax=b\). The above initial-value problem may be viewed as an equation arising from a waveform Newton method whose iterates are to be computed by inner iterations.
- Convolution quadrature and discretized operational calculus. I
- scientific article; zbMATH DE number 3720254 (Why is no real title available?)
- Linear acceleration of Picard-Lindelöf iteration
- Multi-grid dynamic iteration for parabolic equations
- Sets of convergence and stability regions
- Waveform Iteration and the Shifted Picard Splitting
- Pseudospectra of wave from relaxation operators
- Preconditioning waveform relaxation iterations for differential systems
- Waveform Iteration and the Shifted Picard Splitting
- Polynomial acceleration of the Picard-Lindelof iteration
- Chebyshev acceleration for SOR-like method
- Power bounded prolongations and Picard-Lindelöf iteration
- Linear acceleration of Picard-Lindelöf iteration
This page was built for publication: Chebyshev acceleration of Picard-Lindelöf iteration
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1198984)