Cubic spline on quintic spline method for nonlinear boundary value problems (Q1805280): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
ReferenceBot (talk | contribs)
Changed an Item
 
Property / cites work
 
Property / cites work: Cubic spline on quintic spline method for nonlinear boundary value problems / rank
 
Normal rank
Property / cites work
 
Property / cites work: A boundary element method based solution procedure for nonlinear boundary value problems / rank
 
Normal rank
Property / cites work
 
Property / cites work: Solution of two‐point boundary value problems using splines / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5540873 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On the use of splines for the numerical solution of nonlinear two-point boundary value problems / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5652137 / rank
 
Normal rank
Property / cites work
 
Property / cites work: A desirable form for sparse matrices when computing their inverse in factored forms / rank
 
Normal rank

Latest revision as of 14:18, 23 May 2024

scientific article
Language Label Description Also known as
English
Cubic spline on quintic spline method for nonlinear boundary value problems
scientific article

    Statements

    Cubic spline on quintic spline method for nonlinear boundary value problems (English)
    0 references
    11 December 1995
    0 references
    The problem considered is the scalar second-order equation \(d^2 c/dx^2=f(c,x)\) (with boundary conditions). The problem is recast as a 2-dimensional first-order boundary value problem and then treated numerically componentwise via splines over a uniform grid with mesh size \(h=1/n\) and \(x_i=ih\) for \(i=0, \dots, n\). Integrating the boundary value problem yields \[ Q_i(c) := c_i-c_{i-1}-\int^{x_i}_{x_{i-1}} f(x,c(x))dx=0\text{ for }i=0,\dots, n. \] The evaluation of the function \(Q(c)\) is performed via the Euler-Maclaurin formula: \(Q_i (c)=c_i -\) \(c_{i-1}-(h/2) (f_i-f_{i-1})+(h^2/12) (f^1_i-f^1_{i-1})-(h^4/720) (f^3_i-f^3_{i- 1})+(h^6/30240) (f^5_i-f^5_{i-1})+O(h^8)\). To determine \(f^1\) and \(f^3\) from \(f\), the author uses a quintic spline \(S(x)\) to interpolate \(f\), and then \(f^5\) is determined by a cubic spline \(\widehat {S} (x)\) using \(f^3\). This finally yields \(Q_i (c)=c_i-c_{i-1}-\) \((h/2) (f_i-f_{i-1})+(h^2/12) (S^1_i-S^1_{i-1})-(h^4/ 720) (S^3_i-S^3_{i-1})+(h^6/ 30240) (\widehat {S}^5_i-\widehat {S}^5_{i-1})+O(h^8)\). The nonlinear system which is numerically solved is \(u_i(c)+v_i(c)=0\) where \(u_i(c) := c_i-c_{i-1}-(h/2) (f_i-f_{i-1})\) and \(v_i(c) := (h^2/ 12) (S^1_i-S^1_{i-1})-(h^4/ 720) (S^3_i-S^3_{i-1})+(h^6/ 30240) (\widehat {S}^5_i-\widehat {S}^5_{i-1}).\) The system is solved via the modified Newton scheme of solving \((\partial u/\partial c)^{-1} (c^{(k)}) w=-Q(c^{(k)})\) and setting \(c^{(k+1)} := c^{(k)}+w\). Two numerical examples (one of which is nonlinear) are given with boundary conditions \(c'(0)=0\) and \(c(1)=0\).
    0 references
    second-order equation
    0 references
    uniform grid
    0 references
    Euler-Maclaurin formula
    0 references
    quintic spline
    0 references
    cubic spline
    0 references
    Newton scheme
    0 references
    numerical examples
    0 references

    Identifiers