Numerical Verification Methods and Computer-Assisted Proofs for Partial Differential Equations
Numerical Verification Methods and Computer-Assisted Proofs for Partial Differential Equations
复制标题
DOI:
10.1007/978-981-13-7669-6
复制
发表时间:
2019
期刊:
影响因子:
--
通讯作者:
M. Nakao;M. Plum;Yoshitaka Watanabe
中科院分区:
文献类型:
--
作者:
M. Nakao;M. Plum;Yoshitaka Watanabe
This book is a compilation of research studies the authors undertook over many years. The overall goal of the presented methods is to gain information, in particular existence statements and explicit a posteriori bounds, for true solutions to partial differential equation problems, using a combination of analytical arguments and numerical computations. The numerical part includes the computation of an approximate solution, with the aid of which an existence statement and explicit rigorous error bounds for a true solution are derived by analytical fixed-point arguments.This methodological basis gives rise to two slightly different views of such an approach: on one hand, it is an extension of “classical” numerical methods, since it adds rigorous and explicit error bounds to these methods and to the usual asymptotic error estimates. Under this view, the name “verified numerical computation” or “numerical verification” often appears in the corresponding literature. On the other hand, the approach can also be viewed as a part of an analysis, since it provides existence results for partial differential equation problems, including various examples where purely analytical methods have not been successful. With this view in mind, the name “computer-assisted proofs” has often been given to such methods.