Convergence of adaptive BEM for some mixed boundary value problem (Q412294): Difference between revisions

The authors consider the symmetric boundary integral formulation for a 2D mixed (Dirichlet and Neumann) boundary value problem attached to the Laplace equation. The data-perturbed Galerkin formulation is used in order to solve the integral equation. An a posteriori error estimator is introduced. This is a sum of error estimators which control the discretization error and certain data oscillation terms which control the consistency errors introduced by the data approximation. The convergence to zero of the error estimators for the sequence of the discrete solutions produced by the adaptive scheme is proved. Some numerical examples are carried out using the MATLAB implementation of the proposed estimator. They underline the efficiency of the proposed adaptive algorithm.