Lidstone-Euler interpolation and related high even order boundary value problem (Q2044087)

This paper discusses in both theoretical and computational aspects the solvability of the following Lidstone-Euler boundary value problem (LEbvp) \[(-1)^ry^{(2r)}(x)=f(x, y, y',\ldots,y^{(q)}),\] \[y^{(2i)}(0)=\alpha_i,\quad y^{(2i+1)}(1)=\beta_i,\quad i=0,\ldots,r-1,\] where \(0\leq q\leq 2r-1\) is fixed, \(\alpha_i, \beta_i\), \(i=0,\ldots,r-1\), are finite real constants. The authors establish firstly that this problem has a unique solution in \(D\) under the assumption that \(k_i>0\), \(0\leq i \le q\), are real given numbers such that \[\frac{Q}{2^{r-i}}\leq k_{2i},\; i=0,\ldots,\Bigl[\frac{q}{2}\Bigr],\quad \frac{Q}{2^{r-i-1}}\leq k_{2i+1},\; i=0,\ldots,\Bigl[\frac{q-1}{2}\Bigr],\] where \(Q\) is the maximum of \(|f(x,y_0,\ldots,y_q)|\) on the compact \([0,1]\times D\) and \(D=\{(y_0,\ldots,y_q):|y_i|\leq2k_i,\,i=0,\ldots,q\},\) \[C_{2i}<k_{2i}, \; i=0,\ldots,\Bigl[\frac{q}{2}\Bigr],\quad C_{2i+1}<k_{2i+1}, \; i=0,\ldots,\Bigl[\frac{q-1}{2}\Bigr],\] where \[C_{2i}=\frac{2}{3}\sum_{k=0}^{r-i-1}\frac{2^{2k}}{(2k)!\pi^{2k-1}}\Bigl[\frac{2|f^{(2(k+i)+1)}(1)}{(2k+1)\pi}+|f^{(2(k+i))}(0)|\Bigr]\] and \[C_{2i+1}=|f^{(2i+1)}(1)+\frac{2}{3}\sum_{k=1}^{r-i-1}\frac{2^{2k-1}}{(2k-1)!\pi^{2(k-1)}}\Bigl[\frac{|f^{(2(k+i)+1)}(1)}{k\pi}+|f^{(2(k+i))}(0)|\Bigr]\] and \[|f(x,y_0,\ldots,y_q) - f(x,z_0,\ldots,z_q)|<\sum_{i=0}^qL_i|y_i - z_i|,\] where \(L_i>0\) with \(\sum_{i=0}^q\frac{L_i}{2^r}<1\). The rest is summarized by the authors at the end of the paper: ``We introduced a method for the numerical solution of the LEbvp. This method uses the extrapolated Bernstein polynomials. Hence we gave an approximating, convergent polynomial sequence for the numerical solution of the LEbvp. An algorithm for effective calculation is given too. Numerical examples support theoretical results and show that high accuracy in the approximation is achieved by using extrapolation, both in theoretical and computational aspects. A comparison with a modified decomposition method is given for a tenth-order problem''.