A kernel function based regularized method for boundary value problems with noisy information

Abstract

Taking advantage of the reproducing kernel theory, several effective numerical algorithms have been developed to solve boundary value problems (BVPs) with the exact right side functions. However, these methods have difficulty in solving effectively linear boundary value problems when the right side of the equation has contaminated data. The objective of this letter is to introduce a robust numerical algorithm for linear BVPs with noisy right-hand side functions information. To overcome the challenges of the noisy right-hand side functions, the idea of regularization is used. Numerical simulation is employed to illustrate the superiority of the present method.