Identification of the reaction coefficient in time fractional diffusion equations

作者:

Highlights:

摘要

In this paper, we present an inverse problem of identifying the reaction coefficient for time fractional diffusion equations in two dimensional spaces by using boundary Neumann data. It is proved that the forward operator is continuous with respect to the unknown parameter. Because the inverse problem is often ill-posed, regularization strategies are imposed on the least fit-to-data functional to overcome the stability issue. There may exist various kinds of functions to reconstruct. It is crucial to choose a suitable regularization method. We present a multi-parameter regularization L2+BV method for the inverse problem. This can extend the applicability for reconstructing the unknown functions. Rigorous analysis is carried out for the inverse problem. In particular, we analyze the existence and stability of regularized variational problem and the convergence. To reduce the dimension in the inversion for numerical simulation, the unknown coefficient is represented by a suitable set of basis functions based on a priori information. A few numerical examples are presented for the inverse problem in time fractional diffusion equations to confirm the theoretic analysis and the efficacy of the different regularization methods.

论文关键词:Time fractional diffusion equation,Reaction inversion,Multi-parameter regularization

论文评审过程:Received 28 February 2018, Revised 13 June 2018, Available online 2 July 2018, Version of Record 14 July 2018.

论文官网地址:https://doi.org/10.1016/j.cam.2018.06.047