An exponential-trigonometric optimal interpolation formula

The authors present in this paper a derivation of the optimal interpolation formula in Hilbert space \(W_{2}^{(3,0)}(0,1)\) by Sobolev method. Precisely, the following interpolation formula is considered\N\[\N\varphi(x) \tilde{=} P_{\varphi}(x) = \sum\limits_{\beta=0}^{N} C_{\beta}(x)\varphi(x_{\beta}),\N\]\Nwhich satisfies the interpolation conditions \(\varphi(x_{\beta}) = P_{\beta}(x_{\beta})\), \(\beta=0,1,\ldots, N\). Here \(C_{\beta}(x)\) and \(x_{\beta}\in[0,1]\) are called coefficients and nodes of the interpolation formula, respectively. Further, \(W_{2}^{(3,0)}(0,1)\) denotes the set of all functions \(\varphi\) defined on \([0,1]\) which posses an absolutely continuous second derivative on \([0,1]\) and whose third derivative is in \(L_{2}(0,1)\). The set \(W_{2}^{(3,0)}(0,1)\) under the pseudo-inner product\N\[\N\langle \varphi,\psi\rangle = \int\limits_{0}^{1} \left(\varphi^{(3)}(x)+\varphi(x)\right)\left(\psi^{(3)}(x)+\psi(x)\right) dx\N\]\Nis a Hilbert space. The norm is introduced as follows\N\[\N\|\varphi\|_{W_{2}^{(3,0)}(0,1)} = \left\{\int\limits_{0}^{1} \left(\varphi^{(3)}(x)+\varphi(x)\right)^{2} dx\right\}^{\frac{1}{2}}.\N\]\NFor a fixed \(z\in[0,1]\), the error\N\[\N(l,\varphi) = \varphi(z) - P_{\varphi}(z)\N\]\Nof the interpolation formula is a linear functional. Here, for a fixed \(z\in[0,1]\)\N\[\Nl(x,z) = \delta(x-z) - \sum\limits_{\beta=0}^{N} C_{\beta}(z)\delta(x-x_{\beta}),\N\]\Nand the error functional of the interpolation formula belongs to the dual space \(W_{2}^{(3,0)*}(0,1)\).\par After introducing the basic ideas and preliminaries, the authors address two main problems:\N\begin{itemize}\N\item \textbf{Problem 1}. Find the norm of the error functional \(l\) of interpolation formula in the space \(W_{2}^{(3,0)*}(0,1)\).\N\item \textbf{Problem 2}. For fixed interpolation nodes \(x_{\beta}\), find the optimal interpolation coefficients \(\mathring{C}_{\beta}(z)\).\N\end{itemize}\NAdditionally, after introducing some extra conditions, the authors also formulate the third problem, related to the \textbf{Problem 2}: interpolation coefficients \(C_{\beta}(z)\), considered as discrete functions, must satisfy additional constrains formulated in the paper. The fourth problem formulated on this way is related to some technical aspects of the study.

Recommendations

Cites work

Cited in

(3)

This page was built for publication: An exponential-trigonometric optimal interpolation formula

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6544173)