Journal of Computational and Applied Mathematics 298 (2016) 40–52
Contents lists available at ScienceDirect
Journal of Computational and Applied
Mathematics
journal homepage: www.elsevier.com/locate/cam
A new iteratively total variational regularization for
nonlinear inverse problems
✩
Li Li
∗
, Bo Han
Department of mathematics, Harbin Institute of Technology, Harbin, 150001, PR China
a r t i c l e i n f o
Article history:
Received 23 January 2015
Received in revised form 6 September 2015
Keywords:
Ill-posed problems
Total variational regularization
Bregman distance
Homotopy perturbation
a b s t r a c t
Superior to the other well-known iterations, such as Landweber iteration, total variational
regularization is a well-established method for identifying the piecewise smooth param-
eter and processing the image. Therefore, we propose a new iteratively total variational
regularization based on total variational regularization and homotopy perturbation tech-
nique in this paper. Taking advantage of Bregman distance, we construct a nested iteration
to inverse parameters. Meanwhile we prove that the error of the new method decreases
monotonically, and the new iterative method with the noisy data is regular according to
the discrepancy principle. In the last numerical section, compared with Landweber type
total variational regularization and Runge–Kutta type regularization, numerical results of
this new method indicate that this new regularization is time saving for the same accuracy,
and more robust to the noisy data.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
Considering the nonlinear problems
F(u) = y (1.1)
where F : D (F)(⊂ X) → Y , and X, Y are Banach spaces. The exact data on the right-hand side y ∈ range(F), which
guarantees that there is a solution to Eq. (1.1) at least. However, we usually cannot obtain the exact data y, the noisy data y
δ
satisfies ∥y
δ
− y∥ ≤ δ, where δ is the error level. We assume u
∗
is the solution to Eq. (1.1), and the solution does not need to
be unique. Since the ill-posedness of the nonlinear inverse problems (1.1), we have to look for some effective regularization
methods to inverse the problems (1.1) and to find out the approximate solution to the true solution.
The well-known regularization for finding an approximate solution is Tikhonov regularization [1,2]. A minimizer u
δ
α
of
the following Tikhonov-functional
Φ(u) = ∥F (u) − y
δ
∥
2
+ α∥u − u
∗
∥
2
(1.2)
is solved to approximate the solution to problem (1.1), where α is regarded as the regularized parameter and u
δ
α
is the
Tikhonov regularization solution. Moreover, iterative methods, such as Landweber iteration [3–5], Levenberg–Marquardt
iteration [6], and Runge–Kutta type iteration [7,8], are all effective methods for solving ill-posed problems. Assume there
✩
This work is supported by the NSF of China (No. 61307023, No. 41474102), and the Postdoctoral Scientific Research Developmental Fund (No. LBH-
Q14073).
∗
Corresponding author.
E-mail address: lily@hit.edu.cn (L. Li).
http://dx.doi.org/10.1016/j.cam.2015.11.033
0377-0427/© 2015 Elsevier B.V. All rights reserved.