Core Concepts
The paper presents the first formulation of the optimal polynomial approximation of the solution of linear non-autonomous systems of ODEs in the framework of the so-called ⋆-product, and derives upper bounds for its error.
Abstract
The paper introduces the ⋆-product framework for solving linear non-autonomous ordinary differential equations (ODEs). It shows how to formally state the problem of finding the best polynomial approximation of the solution in this framework and derives upper bounds for the approximation error.
Key highlights:
The ⋆-product is the basis of new approaches for the analytical and numerical solution of linear non-autonomous ODEs.
The paper formulates the problem of finding the best polynomial approximation of the solution in the ⋆-framework.
It is shown that the error of the L2-best ⋆-polynomial approximant can be bounded by the best uniform norm polynomial approximation error for the exponential function.
This result is crucial for understanding the numerical behavior of polynomial-based methods when solving linear systems derived using the ⋆-approach.
The paper extends matrix analysis results to the ⋆-framework, including the definition of a ⋆-inner product, ⋆-norm, and induced matrix ⋆-norm.
The main result provides an error bound for the best ⋆-polynomial approximation of the ⋆-resolvent, which is the key to solving the original non-autonomous ODE.