Error estimation based on locally weighted defect for boundary value problems in second order ordinary differential equations (Q486696)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Error estimation based on locally weighted defect for boundary value problems in second order ordinary differential equations
scientific article

    Statements

    Error estimation based on locally weighted defect for boundary value problems in second order ordinary differential equations (English)
    0 references
    0 references
    0 references
    0 references
    16 January 2015
    0 references
    The efficient asymptotical correctness of the a posteriori error estimate is investigated in the numerical approach by piecewise polynomial collocation methods of two-point boundary value problems for second-order ordinary differential equations. For two-point boundary value conditions of Dirichlet type associated to linear and nonlinear second-order ordinary differential equations it is proved that, in the application of a collocation method with piecewise polynomial function of degree \(m+1\), the deviation of the error estimate has the order \(O(h^{(m+2)})\). In this way, the authors prove the high asymptotical correctness of such an error indicator, based on the defect of the collocation solution with respect to an associated exact difference scheme. The way to extend these theoretical results to the case of general linear boundary conditions is presented. The theoretical results are illustrated and confirmed on two numerical experiments, one nonlinear with Dirichlet-type boundary value conditions, and the other one linear with general linear boundary value conditions. Finally, some ideas for the extension of the obtained theoretical results to general nonlinear boundary value conditions are pointed out.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    second-order boundary value problems
    0 references
    collocation methods
    0 references
    asymptotically correct a posteriori error estimates
    0 references
    defect correction principle
    0 references
    exact difference scheme
    0 references
    superconvergence
    0 references
    linear
    0 references
    nonlinear
    0 references
    numerical experiment
    0 references
    0 references
    0 references