Stable Algorithm for Identification of Source Based on Dynamical Systems Methods

Abstract

The problem of source identification is ill-posed in the sense of hadamard, this means that small errors in the measurement may cause very large errors in the numerical result. In this paper, an inverse problem of source identification in heat conduction equations based on Dynamical Systems Methods (DMS) is proposed, and competes with classic Tikhonov regularization algorithms(TV), The numerical results demonstrate the algorithm based on Dynamical Systems Methods is stable and accurate.