NUMERICAL ALGORITHM FOR SOLVING THE INVERSE PROBLEM OF SOURCE TERM IDENTIFICATION OF THE SUBDIFFUSION DIFFERENTIAL EQUATION UNDER STURM TYPE BOUNDARY CONDITIONS

Abstract

This work is devoted to the inverse problem of source term identification for the subdiffusion differential equation under Sturm-type boundary conditions. Additional information in the form of the final overdetermination condition is given. We construct a numerical algorithm for solving the inverse problem. The algorithm is based on the biconjugate gradient stabilized iterative method and the Tikhonov regularization method. To solve the forward initial boundary value subproblems at each iteration, we apply the finite difference scheme. We present the results of numerical experiments that confirm the ability of the developed algorithm to solve the inverse problem in the case of perturbations in input data.