No. of Iter. | Time | |||
n | Alg. 3.1 | Alg. 4.1 | Alg. 3.1 | Alg. 4.1 |
2 | 120 | 154 | 0.288s | 1.133s |
10 | 156 | 201 | 0.691s | 2.516s |
50 | 157 | 202 | 4.641s | 13.853s |
100 | 157 | 203 | 12.333s | 53.145s |
In this study, we present a numerical scheme for solving nonlinear ordinary differential equations with classical and Caputo–Fabrizio derivatives using consecutive interval division and the midpoint approach. By doing so, we increased the accuracy of the midpoint approach, which is dependent on the number of interval divisions. In the example of the Caputo–Fabrizio differential operator, we established the existence and uniqueness of the solution using the Caratheodory-Tonelli sequence. We solved numerous nonlinear equations and determined the global error to test the accuracy of the proposed scheme. When the differential equation met the circumstances under which it was generated, the results revealed that the procedure was quite accurate.
Citation: Abdon Atangana, Seda İğret Araz. A successive midpoint method for nonlinear differential equations with classical and Caputo-Fabrizio derivatives[J]. AIMS Mathematics, 2023, 8(11): 27309-27327. doi: 10.3934/math.20231397
[1] | Xinfeng Liang, Mengya Zhang . Triangular algebras with nonlinear higher Lie n-derivation by local actions. AIMS Mathematics, 2024, 9(2): 2549-2583. doi: 10.3934/math.2024126 |
[2] | He Yuan, Zhuo Liu . Lie n-centralizers of generalized matrix algebras. AIMS Mathematics, 2023, 8(6): 14609-14622. doi: 10.3934/math.2023747 |
[3] | Shan Li, Kaijia Luo, Jiankui Li . Generalized Lie n-derivations on generalized matrix algebras. AIMS Mathematics, 2024, 9(10): 29386-29403. doi: 10.3934/math.20241424 |
[4] | Guangyu An, Xueli Zhang, Jun He, Wenhua Qian . Characterizations of local Lie derivations on von Neumann algebras. AIMS Mathematics, 2022, 7(5): 7519-7527. doi: 10.3934/math.2022422 |
[5] | Junyuan Huang, Xueqing Chen, Zhiqi Chen, Ming Ding . On a conjecture on transposed Poisson n-Lie algebras. AIMS Mathematics, 2024, 9(3): 6709-6733. doi: 10.3934/math.2024327 |
[6] | Baiying He, Siyu Gao . The nonisospectral integrable hierarchies of three generalized Lie algebras. AIMS Mathematics, 2024, 9(10): 27361-27387. doi: 10.3934/math.20241329 |
[7] | Mohd Arif Raza, Huda Eid Almehmadi . Lie (Jordan) σ−centralizer at the zero products on generalized matrix algebra. AIMS Mathematics, 2024, 9(10): 26631-26648. doi: 10.3934/math.20241295 |
[8] | He Yuan, Qian Zhang, Zhendi Gu . Characterizations of generalized Lie n-higher derivations on certain triangular algebras. AIMS Mathematics, 2024, 9(11): 29916-29941. doi: 10.3934/math.20241446 |
[9] | Huizhang Yang, Wei Liu, Yunmei Zhao . Lie symmetry reductions and exact solutions to a generalized two-component Hunter-Saxton system. AIMS Mathematics, 2021, 6(2): 1087-1100. doi: 10.3934/math.2021065 |
[10] | Senrong Xu, Wei Wang, Jia Zhao . Twisted Rota-Baxter operators on Hom-Lie algebras. AIMS Mathematics, 2024, 9(2): 2619-2640. doi: 10.3934/math.2024129 |
In this study, we present a numerical scheme for solving nonlinear ordinary differential equations with classical and Caputo–Fabrizio derivatives using consecutive interval division and the midpoint approach. By doing so, we increased the accuracy of the midpoint approach, which is dependent on the number of interval divisions. In the example of the Caputo–Fabrizio differential operator, we established the existence and uniqueness of the solution using the Caratheodory-Tonelli sequence. We solved numerous nonlinear equations and determined the global error to test the accuracy of the proposed scheme. When the differential equation met the circumstances under which it was generated, the results revealed that the procedure was quite accurate.
The variational inequality problem (VIP) was introduced by Stampacchia [1] and provided a very useful tool for researching a large variety of interesting problems arising in physics, economics, finance, elasticity, optimization, network analysis, medical images, water resources, and structural analysis, see for example ([2,3,4,5,6,7,8,9,10,11,12,13,14,15]) and references therein.
Let H be a real Hilbert space with inner product ⟨⋅,⋅⟩ and norm ‖⋅‖, respectively. Let C be a nonempty closed convex subset of H. Let B:C⟶H be an operator.
In this article, our study is related to a classical variational inequality problem (VIP) which aims to find an element x†∈C such that
⟨Bx†,x−x†⟩≥0, ∀x∈C. | (1.1) |
It is well known that x♯∈VI(B,C) if and only if x♯=PC(x♯−ζBx♯), where ζ>0, in other words, the VIP is equivalent to the fixed point problem (see [16]). Supposing that B is η-strongly monotone and L-Lipschitz continuous with 0<ζ<2ηL2, the following sequence {xn} of Picard iterates:
xn+1=PC(xn−ζBxn), | (1.2) |
converges strongly to a point x†∈VI(B,C) due to the fact that PC(I−ζB) is a contraction on C. However, in general, the algorithm (1.2) fails when B is monotone and L-Lipschitz continuous (see [17]). In [7], Korpelevich put forward an extragradient method which provided an important idea for solving monotone variational inequality:
yn=PC(xn−λfxn),xn+1=PC(xn−λfyn), | (1.3) |
where f is monotone, L-Lipschitz continuous in the finite dimensional Euclidean space Rn and λ∈(0,1L).
The another motivation of this article is the split common fixed point problem which aims to find a point u∈H1 such that
u∈Fix(T) and Au∈Fix(S). | (1.4) |
The split common fixed point problem can be regarded as a generalization of the split feasibility problem. Recall that the split feasibility problem is to find a point satisfying
u∈C and Au∈Q, | (1.5) |
where C and Q are two nonempty closed convex subsets of real Hilbert spaces H1 and H2, respectively and A:H1→H2 is a bounded linear operator. Inverse problems in various disciplines can be expressed as the split feasibility problem and the split common fixed point problem. Problem (1.4) was firstly introduced by Censor and Segal [18]. Note that solving (1.4) can be translated to solve the fixed point equation:
u=S(u−τA∗(I−T)Au), τ>0. |
Whereafter, Censor and Segal proposed an algorithm for directed operators. Since then, there has been growing interest in the split common fixed point problem (see [19,20,21,22]).
Censor et al. [23] first proposed split variational inequality problems by combining the variational inequality problem and the split feasibility problem. Very recently, in 2017, Tian and Jiang [24] considered the following split variational inequality problem: finding an element u such that
u∈VI(A,C) and Bu∈Fix(T), | (1.6) |
where T:H2→H2 is nonexpansive, B:H1→H2 is a bounded linear operator with its adjoint B∗, and A:C→H1 is a monotone and L-Lipschitz continuous mapping. Then they presented the following iteration method by combining the extragradient method with CQ algorithm for solving the (1.6):
Algorithm 1.1. Choose an arbitrary initial value x1∈C. Assume xn has been constructed. Compute
yn=PC(xn−τnA∗(I−T)Axn),zn=PC(yn−ςnFyn),xn+1=PC(yn−ςnFzn). | (1.7) |
They proved that the iterative sequence {xn} defined by Eq (1.7) converges weakly to an element z∈Γ, where Γ is the set of solutions of the problem (1.6). However, Algorithm 1.1 fails, in general, to converge strongly in the setting of infinite-dimensional Hilbert spaces. We also notice that Algorithm 1.1 is involved with three metric projections in each iteration, which might seriously affect the efficiency of the method.
Motivated and inspired by the above works, in the present paper, we consider variational inequality problems and split common fixed point problems for finding an element u such that
ˆx∈VI(A,C) and Bˆx∈∞⋂n=1Fix(Tn), | (1.8) |
where {Tn}∞n=1:H2→H2 is an infinite family of nonexpansive mappings, B:H1→H2 is a bounded linear operator with its adjoint B∗, and A:H1→H1 is a monotone and L-Lipschitz continuous mapping. In contrast to Tian and Jiang [24], we consider the common fixed points of an infinite family of nonexpansive mappings instead of only the fixed points of a nonexpansive mapping. The efficiency of the algorithm is also improved by removing the projection operator in the first iteration which might affect the efficiency of the method to a certain extent. Finally, we present a very simple modification to extragradient method, which makes our algorithm have the strong convergence. It is well known that the strong convergence theorem is always more convenient to use.
This paper is organized as follows: In Section 2, we give some definitions and key lemmas which are used in this paper. Section 3 consists of our algorithms and provides the strong convergence theorems. In Section 4, numerical examples are provided for illustration. Finally, this paper is concluded in Section 5.
Let H be a real Hilbert space with inner product ⟨⋅,⋅⟩ and norm ‖⋅‖, respectively. Let C be a nonempty closed convex subset of H. Let T:C⟶C be an operator. We use Fix(T) to denote the set of fixed points of T, that is, Fix(T)={x†|x†=Tx†,x†∈C}. First, we give some definitions and lemmas related to the involved operators.
Definition 2.1. An operator T:C⟶C is said to be nonexpansive if ‖Tu−Tv‖≤‖u−v‖ for all u,v∈C.
Definition 2.2. An operator A:C⟶H is said to be monotone if ⟨Ax−Ay,x−y⟩≥0 for all x,y∈C.
A monotone operator R:H⇉2H is called maximal monotone if the graph of R is a maximal monotone set.
Definition 2.3. An operator T:C⟶H is said to be L-Lipschitzian if there exists L>0 such that ‖Tx−Ty‖≤L‖x−y‖ for all x,y∈C.
Usually, the convergence of fixed point algorithms requires some additional smoothness properties of the mapping T such as demi-closedness.
Definition 2.4. An operator T is said to be demiclosed if, for any sequence {un} which weakly converges to u∗, and if Tun⟶w, then Tu∗=w.
Recall that the (nearest point or metric) projection from H onto C, denoted by PC, assigns to each u∈H, the unique point PCu∈C with the property:
‖u−PCu‖=inf |
The metric projection \mathcal{P}_{\mathcal{C}} of \mathcal{H} onto \mathcal{C} is characterized by
\begin{eqnarray} \begin{split} &\langle u\;-\mathcal{P}_{\mathcal{C}}u,v-\mathcal{P}_{\mathcal{C}}u\rangle\leq0\\ \; \mathrm{or}&\ \|u-v\|^{2}\geq\|u-\mathcal{P}_{\mathcal{C}}u\|^{2}+\|v-\mathcal{P}_{\mathcal{C}}u\|^{2} \end{split} \end{eqnarray} | (2.1) |
for all u\in \mathcal{H}, v\in \mathcal{C} . It is well known that the metric projection \mathcal{P}_{\mathcal{C}}:\mathcal{H}\rightarrow \mathcal{C} is firmly nonexpansive, that is,
\begin{eqnarray} \begin{split} &\langle u\;-v,\mathcal{P}_{\mathcal{C}}u-\mathcal{P}_{\mathcal{C}}v\rangle\geq\|\mathcal{P}_{\mathcal{C}}u-\mathcal{P}_{\mathcal{C}}v\|^{2}\\ \; \mathrm{or}&\ \|\mathcal{P}_{\mathcal{C}}u-\mathcal{P}_{\mathcal{C}}v\|^{2}\leq\|u-v\|^{2}-\|(I-\mathcal{P}_{\mathcal{C}})u-(I-\mathcal{P}_{\mathcal{C}})v\|^{2} \end{split} \end{eqnarray} | (2.2) |
for all u, v\in \mathcal{H} . More information on the metric projection can be found, for example, in Section 3 of the book by Goebel et al. (see [25]).
For all u, v \in H , the following conclusions hold:
\begin{eqnarray} \|tu+(1-t)v\|^{2} = t\|u\|^{2}+(1-t)\|v\|^{2}-t(1-t)\|u-v\|^{2},\ \ t\in[0,1], \end{eqnarray} | (2.3) |
\begin{eqnarray} \|u+v\|^{2} = \|u\|^{2}+2\langle u,v\rangle+\|v\|^{2} \end{eqnarray} | (2.4) |
and
\begin{eqnarray} \|u+v\|^{2}\leq\|u\|^{2}+2\langle v,u+v\rangle. \end{eqnarray} | (2.5) |
Let \{\mathcal{T}_{n}\}^{\infty}_{n = 1}:\mathcal{H}\rightarrow \mathcal{H} be an infinite family of nonexpansive mappings and \lambda_{1}, \lambda_{2}, ... be real numbers such that 0\leq\lambda_{i}\leq1 for each i\in\mathbb{N} . For any n\in\mathbb{N} , define a mapping \mathcal{W}_{n} of \mathcal{C} into \mathcal{H} as follows:
\begin{eqnarray} \begin{split} \;&\mathcal{U}_{n,n+1} = \mathcal{I},\\ \;&\mathcal{U}_{n,n} = \lambda_{n}\mathcal{T}_{n}\mathcal{U}_{n,n+1}+ (1-\lambda_{n})\mathcal{I},\\ \;&\mathcal{U}_{n,n-1} = \lambda_{n-1}\mathcal{T}_{n-1}\mathcal{U}_{n,n}+ (1-\lambda_{n-1})\mathcal{I},\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \\ \;&\ldots\\ \;&\mathcal{U}_{n,k} = \lambda_{k}\mathcal{T}_{k}\mathcal{U}_{n,k+1}+ (1-\lambda_{k})\mathcal{I},\\ \;&\mathcal{U}_{n,k-1} = \lambda_{k-1}\mathcal{T}_{k-1}\mathcal{U}_{n,k}+ (1-\lambda_{k-1})\mathcal{I},\\ \;&\ldots\\ \;&\mathcal{U}_{n,2} = \lambda_{2}\mathcal{T}_{2}\mathcal{U}_{n,3}+ (1-\lambda_{2})\mathcal{I},\\ \;&\mathcal{W}_{n} = \mathcal{U}_{n,1} = \lambda_{1}\mathcal{T}_{1}\mathcal{U}_{n,2}+ (1-\lambda_{1})\mathcal{I}.\\ \end{split} \end{eqnarray} | (2.6) |
Such a mapping \mathcal{W}_{n} is called the \mathcal{W} -mapping generated by \mathcal{T}_{1}, \mathcal{T}_{2}, ..., \mathcal{T}_{n} and \lambda_{1}, \lambda_{2}, ..., \lambda_{n} . We have the following crucial Lemma concerning \mathcal{W}_{n} :
Lemma 2.1. [26] Let \mathcal{H} be a real Hilbert space. Let \{\mathcal{T}_{n}\}^{\infty}_{n = 1}:\mathcal{H}\rightarrow \mathcal{H} be an infinite family of nonexpansive mappings such that \bigcap^{\infty}_{n = 1}\mathrm{Fix} (\mathcal{T}_{n})\neq\emptyset . Let \lambda_{1}, \lambda_{2}, ... be real numbers such that 0\leq\lambda_{i}\leq b < 1 for each i\geq1 . Then we have the following:
(1) For any x\in\mathcal{H} and k\geq1 , the limit \lim_{n\rightarrow \infty}\mathcal{U}_{n, k}x exists;
(2) \mathrm{Fix} (\mathcal{W}) = \bigcap^{\infty}_{n = 1}\mathrm{Fix} (\mathcal{T}_{n}) , where \mathcal{W}x = \lim_{n\rightarrow \infty}\mathcal{W}_{n}x = \lim_{n\rightarrow \infty}\mathcal{U}_{n, 1}x, \ \forall x\in\mathcal{C};
(3) For any bounded sequence \{x_{n}\}\subset \mathcal{H} , \lim_{n\rightarrow \infty}\mathcal{W}x_{n} = \lim_{n\rightarrow \infty}\mathcal{W}_{n}x_{n} .
Lemma 2.2. [27] Assume that \{\alpha_{n}\} is a sequence of nonnegative real numbers such that
\alpha_{n+1}\leq(1-\gamma_{n})\alpha_{n} + \delta_{n}, \ \ n\in \mathbb{N}, |
where \{\gamma_{n}\} is a sequence in (0, 1) and \{\delta_{n}\} is a sequence such that
(1) \sum^{\infty}_{n = 1}\gamma_{n} = \infty ;
(2) \limsup_{n\rightarrow \infty}\frac{\delta_{n}}{\gamma_{n}}\leq0 or \sum^{\infty}_{n = 1}|\delta_{n}| < \infty. Then \lim_{n\rightarrow \infty}\alpha_{n} = 0.
Lemma 2.3. [28] Let \{\varpi_{n}\} be a sequence of real numbers. Assume there exists at least a subsequence \{\varpi_{n_{k}}\} of \{\varpi_{n}\} such that \varpi_{n_{k}}\leq\varpi_{n_{k+1}} for all k\geq0 . For every n\geq N_{0} , define an integer sequence \{\tau(n)\} as:
\tau(n) = \max\{i\leq n:\varpi_{n_{i}} < \varpi_{n_{i+1}}\}. |
Then, \tau(n) \rightarrow \infty as n\rightarrow \infty and for all n\geq N_{0} , we have max\{ \varpi_{\tau(n)}, \varpi_{n}\}\leq\varpi_{\tau(n)+1} .
In this section, we introduce our algorithm and prove its strong convergence. Some assumptions on the underlying spaces and involved operators are listed below.
(R_{1}) \mathcal{H}_{1} and \mathcal{H}_{2} are two real Hilbert spaces and \mathcal{C}\subset\mathcal{H}_{1} is a nonempty closed convex subset.
(R_{2}) \mathcal{B}:\mathcal{H}_{1}\rightarrow \mathcal{H}_{2} is a bounded linear operator with its adjoint \mathcal{B}^{\ast} .
(R_{3}) \mathcal{A}:\mathcal{H}_{1}\rightarrow \mathcal{H}_{1} is a monotone and L -Lipschitz continuous mapping.
(R_{4}) \Omega = \{\hat{x}|\hat{x}\in VI(\mathcal{A}, \mathcal{C}) and \mathcal{B}\hat{x}\in \bigcap^{\infty}_{n = 1} \mathrm{Fix}(\mathcal{T}_{n})\} , where \Omega is the set of solutions of the problem (1.8).
Next, we present the following iterative algorithm to find a point \hat{x}\in\Omega .
Algorithm 3.1. Choose an arbitrary initial value x_{1}\in \mathcal{H} . Assume x_{n} has been constructed. Compute
\begin{eqnarray} \begin{split} \;& y_{n} = x_{n}-\tau_{n} \mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n},\\ \;& z_{n} = \mathcal{P}_{\mathcal{C}}(y_{n}-\varsigma_{n}\mathcal{A}y_{n}),\\ \;&x_{n+1} = \mathcal{P}_{\mathcal{C}}((1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})), \\ \end{split} \end{eqnarray} | (3.1) |
where \{\alpha_{n}\} is a sequence in (0, 1) , \varsigma_{n} is a sequence in (0, \frac{1}{L}) , and \tau_{n} is a sequence in (0, \frac{1}{\|\mathcal{B}\|^{ 2}}) .
Theorem 3.1. If \Omega\neq\emptyset and the following conditions are satisfied:
( C_{1} ) \lim_{n\rightarrow \infty}\alpha_{n} = 0\ and\ \sum^{\infty}_{n = 0}\alpha_{n} = \infty ;
( C_{2} ) 0 < \liminf_{n\rightarrow \infty}\varsigma_{n}\leq \limsup_{n\rightarrow \infty}\varsigma_{n} < \frac{1}{L} ;
( C_{3} ) 0 < \liminf_{n\rightarrow \infty}\tau_{n}\leq \limsup_{n\rightarrow \infty}\tau_{n} < \frac{1}{\|\mathcal{B}\|^{ 2}} .
Then, the iterative sequence \{x_{n}\} defined by Eq (3.1) strongly converges to the minimum-norm solution \hat{x}( = \mathcal{P}_{\Omega}\theta) .
Proof. Set z = \mathcal{P}_{\Omega}\theta . We can obtain that
\begin{eqnarray} \begin{split} &\|y_{n}-z\|^{2}\\ = \;&\|x_{n}-z-\tau_{n} \mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ = \;&\|x_{n}-z\|^{2}-2\tau_{n} \langle x_{n}-z,\mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n}\rangle+\|\tau_{n}\mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ = \;&\|x_{n}-z\|^{2}-2\tau_{n} \langle \mathcal{B}x_{n}-\mathcal{B}z,(I-\mathcal{W}_{n})\mathcal{B}x_{n}\rangle+\|\tau_{n}\mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ \leq\;&\|x_{n}-z\|^{2}-\tau_{n} \|(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}+\tau_{n}^{2}\|\mathcal{B}\|^{2}\cdot\|(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ \leq\;&\|x_{n}-z\|^{2}-\tau_{n}(1-\tau_{n}\|\mathcal{B}\|^{ 2})\|(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ \leq\;&\|x_{n}-z\|^{2}. \end{split} \end{eqnarray} | (3.2) |
It follows from (2.1) that
\begin{eqnarray} \begin{split} &\|x_{n+1}-z\|^{2}\\ = \;&\|\mathcal{P}_{\mathcal{C}}((1-\alpha_{n}) (y_{n}-\varsigma_{n}\mathcal{A}z_{n}))-z\|^{2}\\ \leq\;&\|(1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})-z\|^{2}- \|(1-\alpha_{n})(y_{n} -\varsigma_{n}\mathcal{A}z_{n})-x_{n+1}\|^{2}\\ \leq\;&\|(1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n}-z)+\alpha_{n}(-z)\|^{2} \\ \;&-\|(1-\alpha_{n})(y_{n} -\varsigma_{n}\mathcal{A}z_{n}-x_{n+1})+\alpha_{n}(-x_{n+1})\|^{2}\\ \leq\;&(1-\alpha_{n})\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}-z\|^{2}+\alpha_{n}\|-z\|^{2}\\ \;&-(1-\alpha_{n})\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}\|^{2}\\ \;&-((1-\alpha_{n})\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}-x_{n+1}\|^{2}+\alpha_{n}\|-x_{n+1}\|^{2}\\ \;&-(1-\alpha_{n})\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}\|^{2})\\ = \;&(1-\alpha_{n})(\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}-z\|^{2}-\|y_{n} -\varsigma_{n}\mathcal{A}z_{n}-x_{n+1}\|^{2})\\ \;&+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}). \end{split} \end{eqnarray} | (3.3) |
We also observe that
\begin{eqnarray} \begin{split} &\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}-z\|^{2}-\|y_{n} -\varsigma_{n}\mathcal{A}z_{n}-x_{n+1}\|^{2}\\ = \;& \|y_{n}-z\|^{2}-\|y_{n} -x_{n+1}\|^{2}+2\varsigma_{n}\langle\mathcal{A}z_{n},z-x_{n+1}\rangle\\ = \;&\|y_{n}-z\|^{2}-\|y_{n} -x_{n+1}\|^{2}+2\varsigma_{n}\langle\mathcal{A}z_{n},z-z_{n}\rangle +2\varsigma_{n}\langle\mathcal{A}z_{n},z_{n}-x_{n+1}\rangle\\ = \;&\|y_{n}-z\|^{2}-\|y_{n} -x_{n+1}\|^{2}+2\varsigma_{n}\langle\mathcal{A}z_{n}-\mathcal{A}z,z-z_{n}\rangle\\ \;&+2\varsigma_{n}\langle\mathcal{A}z,z-z_{n}\rangle +2\varsigma_{n}\langle\mathcal{A}z_{n},z_{n}-x_{n+1}\rangle\\ \geq\;&\|y_{n}-z\|^{2}-\|y_{n} -x_{n+1}\|^{2} +2\varsigma_{n}\langle\mathcal{A}z_{n},z_{n}-x_{n+1}\rangle\\ = \;&\|y_{n}-z\|^{2}-\|y_{n} -z_{n}\|^{2}-\|z_{n} -x_{n+1}\|^{2}\\ \;&+2\langle y_{n}-\varsigma_{n} \mathcal{A}z_{n}-z_{n},x_{n+1}-z_{n}\rangle.\\ \end{split} \end{eqnarray} | (3.4) |
On the other hand, we have that
\begin{eqnarray} \begin{split} &\langle y_{n}-\varsigma_{n} \mathcal{A}z_{n}-z_{n},x_{n+1}-z_{n}\rangle\\ = \;& \langle y_{n}-\varsigma_{n} \mathcal{A}y_{n}-z_{n},x_{n+1}-z_{n}\rangle+ \varsigma_{n}\langle \mathcal{A}y_{n}-\mathcal{A}z_{n},x_{n+1}-z_{n}\rangle\\ \leq\;&\varsigma_{n}\langle \mathcal{A}y_{n}-\mathcal{A}z_{n},x_{n+1}-z_{n}\rangle\\ \leq\;&\varsigma_{n}\|\mathcal{A}y_{n}-\mathcal{A}z_{n}\|\times\|x_{n+1}-z_{n}\|\\ \leq\;&\varsigma_{n} L\|y_{n}-z_{n}\|\times\|x_{n+1}-z_{n}\|.\\ \end{split} \end{eqnarray} | (3.5) |
Hence, we can derive that
\begin{eqnarray} \begin{split} \|x_{n+1}-z\|^{2} = \;&(1-\alpha_{n})(\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}-z\|^{2}-\|y_{n} -\varsigma_{n}\mathcal{A}z_{n}-x_{n+1}\|^{2})+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}),\\ (by(3.4))\leq\;&(1-\alpha_{n})(\|y_{n}-z\|^{2}-\|y_{n} -z_{n}\|^{2}-\|z_{n} -x_{n+1}\|^{2}\\ \;&+2\langle y_{n}-\varsigma_{n} \mathcal{A}z_{n}-z_{n},x_{n+1}-z_{n}\rangle)+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}),\\ (by(3.5))\leq\;&(1-\alpha_{n})(\|y_{n}-z\|^{2}-\|y_{n} -z_{n}\|^{2}-\|z_{n} -x_{n+1}\|^{2}\\ \;&+2\varsigma_{n} L\|y_{n}-z_{n}\|\times\|x_{n+1}-z_{n}\|)+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2})\\ \leq\;&(1-\alpha_{n})(\|y_{n}-z\|^{2}-\|y_{n} -z_{n}\|^{2}-\|z_{n} -x_{n+1}\|^{2}\\ \;&+\varsigma_{n}^{2} L^{2}\|y_{n}-z_{n}\|^{2}+\|x_{n+1}-z_{n}\|^{2})+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2})\\ \leq\;&(1-\alpha_{n})(\|y_{n}-z\|^{2}+(\varsigma_{n}^{2} L^{2}-1)\|y_{n}-z_{n}\|^{2})+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}),\\ (by(3.2))\leq\;&(1-\alpha_{n})(\|x_{n}-z\|^{2}+(\varsigma_{n}^{2} L^{2}-1)\|y_{n}-z_{n}\|^{2})+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}).\\ \end{split} \end{eqnarray} | (3.6) |
Owing to the assumption ( C_{2} ), it follows from (3.6) that
\begin{eqnarray} \begin{split} \|x_{n+1}-z\|^{2} \leq\;&(1-\alpha_{n})\|y_{n}-z\|^{2}+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}),\\ (by(3.2))\leq\;&(1-\alpha_{n})\|x_{n}-z\|^{2}+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2})\\ \leq\;&(1-\alpha_{n})\|x_{n}-z\|^{2}+\alpha_{n}\|z\|^{2}\\ \leq\;&\max\{\|x_{n}-z\|^{2},\|z\|^{2}\}\\ \end{split} \end{eqnarray} | (3.7) |
and so
\begin{eqnarray} \begin{split} \|x_{n}-z\|^{2} \leq\max\{\|x_{1}-z\|^{2},\|z\|^{2}\}, \end{split} \end{eqnarray} | (3.8) |
which implies that the sequence \{x_{n}\} is bounded. In view of (3.2) and (3.7), we obtain that
\begin{eqnarray} \begin{split} &\tau_{n}(1-\tau_{n}\|\mathcal{B}\|^{ 2})\|(I-\mathcal{W}_{n})\mathcal{B}x_{n}\|^{2}\\ \leq\;& \|x_{n}-z\|^{2}-\|y_{n}-z\|^{2}\\ \leq\;& \|x_{n}-z\|^{2}-\|x_{n+1}-z\|^{2}+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}-\|y_{n}-z\|^{2}).\\ \end{split} \end{eqnarray} | (3.9) |
CASE \mathbb{I} . Suppose that there exists m > 0 such that the sequence \{\|x_{n}-z\|\} is decreasing when n\geq m . Then, \lim_{n\rightarrow \infty}\|x_{n}-z\| exists. Consequently, according to the assumptions ( C_{1} ) and ( C_{3} ), we deduce that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|(I-\mathcal{W}_{n})\mathcal{B}x_{n}\| = 0. \end{split} \end{eqnarray} | (3.10) |
In virtue of the boundedness of the sequence \{\mathcal{B}x_{n}\} and Lemma 2.1, we get that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|W\mathcal{B}x_{n}- W_{n}\mathcal{B}x_{n}\| = 0. \end{split} \end{eqnarray} | (3.11) |
This together with (3.24) implies that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|(I-\mathcal{W})\mathcal{B}x_{n}\| = 0. \end{split} \end{eqnarray} | (3.12) |
It follows from (3.6) that
\begin{eqnarray} \begin{split} &(1-\alpha_{n})(1-\varsigma_{n}^{2} L^{2})\|y_{n}-z_{n}\|^{2}\\ \leq\;&(1-\alpha_{n})\|x_{n}-z\|^{2}-\|x_{n+1}-z\|^{2}\\ \;&+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2})\\ \leq\;&\|x_{n}-z\|^{2}-\|x_{n+1}-z\|^{2}\\ \;&+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}-\|x_{n}-z\|^{2}).\\ \end{split} \end{eqnarray} | (3.13) |
Thanks to the boundedness of the sequence \{x_{n}\} , we derive that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|y_{n}-z_{n}\| = 0. \end{split} \end{eqnarray} | (3.14) |
In view of (3.30), we can also get that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|y_{n}-x_{n}\| = \lim\limits_{n\rightarrow \infty}\|\tau_{n} \mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n}\| = 0(by(3.10)). \end{split} \end{eqnarray} | (3.15) |
Combining (3.14) and (3.15), we obtain that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|z_{n}-x_{n}\| = 0. \end{split} \end{eqnarray} | (3.16) |
On the other hand, we get that
\begin{eqnarray} \begin{split} \|x_{n+1}-z_{n}\| = \;&\|\mathcal{P}_{\mathcal{C}}((1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n}))-\mathcal{P}_{\mathcal{C}}(y_{n}-\varsigma_{n}\mathcal{A}y_{n})\|\\ \leq\;&\|(1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})-(y_{n}-\varsigma_{n}\mathcal{A}y_{n})\|\\ \leq\;&\|(y_{n}-\varsigma_{n}\mathcal{A}z_{n})-(y_{n}-\varsigma_{n}\mathcal{A}y_{n})\|+\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n})\|\\ \leq\;&\|\varsigma_{n}\mathcal{A}z_{n}-\varsigma_{n}\mathcal{A}y_{n}\|+\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}\|\\ \leq\;&\varsigma_{n}\|\mathcal{A}z_{n}-\mathcal{A}y_{n}\| +\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}\|\\ \leq\;&\varsigma_{n}L\|z_{n}-y_{n}\| +\alpha_{n}\|y_{n}-\varsigma_{n}\mathcal{A}z_{n}\|.\\ \end{split} \end{eqnarray} | (3.17) |
Hence, by (3.14), it turns out that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|x_{n+1}-z_{n}\| = 0 \end{split} \end{eqnarray} | (3.18) |
and consequently, according to (3.16), we have that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|x_{n+1}-x_{n}\| = 0. \end{split} \end{eqnarray} | (3.19) |
Next, we can take a subsequence \{n_{i}\} such that
\begin{eqnarray} \begin{split} \limsup\limits_{n\rightarrow \infty}(\|z\|^{2}-\|x_{n+1}\|^{2}) = \;&\lim\limits_{i\rightarrow \infty}(\|z\|^{2}-\|x_{n_{i}+1}\|^{2}).\\ \end{split} \end{eqnarray} | (3.20) |
By the boundedness of the real sequence \{x_{n_{i}+1}\}, we may assume that x_{n_{i}+1}\rightharpoonup x^{\dagger}. Since \mathcal{W} is nonexpansive, we can derive that \mathcal{B}x^{\dagger} = \mathcal{W}\mathcal{B}x^{\dagger} (see Corollary 4.28 in [29]), that is, \mathcal{B}x^{\dagger}\in Fix(\mathcal{W}) = \bigcap^{\infty}_{n = 1}Fix(\mathcal{T}_{n}) .
Now, we show that x^{\dagger}\in VI(\mathcal{A}, \mathcal{C}) . Let
\begin{equation} \mathcal{R}(v) = \left\{ \begin{aligned} &\mathcal{A}v+\mathcal{N}_{\mathcal{C}}(v),\ &v\in \mathcal{C}, \\ &\emptyset\ \ &v\not\in \mathcal{ C}, \end{aligned} \right. \end{equation} | (3.21) |
where \mathcal{N}_{\mathcal{C}}(v) is the normal cone to \mathcal{C} at v . According to Reference [30], we can easily derive that \mathcal{R} is maximal monotone. Let (v, w)\in G(\mathcal{R}) . Since w -Av \in\mathcal{N}_{\mathcal{C}}(v) and x_{n}\in \mathcal{C} , we have that
\langle v-x_{n},w -Av\rangle\geq0. |
Noting that, due to v\in \mathcal{C} , we get
\langle v-x_{n+1},x_{n+1}-(1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})) \rangle\geq0. |
It follows that
\langle v-x_{n+1},\frac{x_{n+1}-y_{n}}{\varsigma_{n}}+\mathcal{A}z_{n}+ \frac{\alpha_{n}}{\varsigma_{n}}(y_{n}-\varsigma_{n}\mathcal{A}z_{n})\rangle\geq0. |
Thus, we can deduce that
\begin{eqnarray} \begin{split} &\langle v-x_{n_{i}+1},w \rangle\\ \geq\;&\langle v-x_{n_{i}+1},Av\rangle\\ \geq\;&-\langle v-x_{n_{i}+1},\frac{x_{n_{i}+1}-y_{n_{i}}}{\varsigma_{n_{i}}}+\mathcal{A}z_{n_{i}}+ \frac{\alpha_{n_{i}}}{\varsigma_{n_{i}}}(y_{n_{i}}-\varsigma_{n_{i}}\mathcal{A}z_{n_{i}})\rangle\\ \;&+\langle v-x_{n_{i}+1},Av\rangle\\ \geq\;&\langle v-x_{n_{i}+1},\mathcal{A}v-\mathcal{A}z_{n_{i}}\rangle-\langle v-x_{n_{i}+1},\frac{x_{n_{i}+1}-y_{n_{i}}}{\varsigma_{n_{i}}} \rangle\\ \;&-\langle v-x_{n_{i}+1}, \frac{\alpha_{n_{i}}}{\varsigma_{n_{i}}}(y_{n_{i}}-\varsigma_{n_{i}}\mathcal{A}z_{n_{i}})\rangle\\ \geq\;&\langle v-x_{n_{i}+1},\mathcal{A}v-\mathcal{A}x_{n_{i}+1}\rangle+ \langle v-x_{n_{i}+1},\mathcal{A}x_{n_{i}+1}-\mathcal{A}z_{n_{i}}\rangle\\ \;&-\langle v-x_{n_{i}+1},\frac{x_{n_{i}+1}-y_{n_{i}}}{\varsigma_{n_{i}}} \rangle- \langle v-x_{n_{i}+1}, \frac{\alpha_{n_{i}}}{\varsigma_{n_{i}}}(y_{n_{i}}-\varsigma_{n_{i}}\mathcal{A}z_{n_{i}})\rangle\\ \geq \;&-\langle v-x_{n_{i}+1},\frac{x_{n_{i}+1}-y_{n_{i}}}{\varsigma_{n_{i}}} \rangle- \langle v-x_{n_{i}+1}, \frac{\alpha_{n_{i}}}{\varsigma_{n_{i}}}(y_{n_{i}}-\varsigma_{n_{i}}\mathcal{A}z_{n_{i}})\rangle\\ \;&+\langle v-x_{n_{i}+1},\mathcal{A}x_{n_{i}+1}-\mathcal{A}z_{n_{i}}\rangle.\\ \end{split} \end{eqnarray} | (3.22) |
As i\rightarrow \infty , we obtain that
\langle v-x^{\dagger},w \rangle\geq0. |
By the maximal monotonicity of \mathcal{R} , we derive that x^{\dagger}\in\mathcal{R}^{-1}0 . Hence, x^{\dagger}\in VI(\mathcal{A}, \mathcal{C}) . Therefore, x^{\dagger}\in\Omega . Since the norm of the Hilbert space \mathcal{H}_{1} is weakly lower semicontinuous(see Lemma 2.42 in [29]), we have the following inequality:
\|x^{\dagger}\|\leq\liminf\limits_{i\rightarrow \infty}\|x_{n_{i}+1}\| |
and therefore
-\|x^{\dagger}\|\geq\limsup\limits_{i\rightarrow \infty}(-\|x_{n_{i}+1}\|). |
From (3.7), we observe that
\begin{eqnarray} \begin{split} \|x_{n+1}-z\|^{2} \leq\;&(1-\alpha_{n})\|x_{n}-z\|^{2}+\alpha_{n}(\|z\|^{2}-\|x_{n+1}\|^{2}).\\ \end{split} \end{eqnarray} | (3.23) |
Thanks to z = \mathcal{P}_{\Omega}\theta and x^{\dagger}\in\Omega , we can deduce that
\limsup\limits_{n\rightarrow \infty}(\|z\|^{2}-\|x_{n_{i}+1}\|^{2}) = \|z\|^{2}+\limsup\limits_{n\rightarrow \infty}(-\|x_{n_{i}+1}\|^{2})\leq\|z\|^{2}-\|x^{\dagger}\|^{2}\leq0. |
Applying Lemma 2.2 to (3.23), we derive that \lim_{n\rightarrow \infty}\|x_{n}-z\| = 0 , which implies that the sequence \{x_{n}\} converges strongly to z .
CASE \mathbb{II} . For any n_{0} , there exists an integer m\geq n_{0} such that \|x_{m}-z\|\leq\|x_{m+1}-z\| . At this case, we set \varpi_{n} = \|x_{n}-z\| . For n\geq n_{0} , we define a sequence \{\tau_{n}\} by
\tau(n) = \max\{l\in\mathbb{N}|n_{0}\leq l\leq n,\varpi_{l}\leq\varpi_{l+1}\}. |
It is easy to show that \tau(n) is a non-decreasing sequence such that
\lim\limits_{n\rightarrow \infty}\tau(n) = +\infty |
and
\varpi_{\tau(n)}\leq\varpi_{\tau(n)+1}. |
This together with (3.9) implies that
\begin{eqnarray} \begin{split} \lim\limits_{n\rightarrow \infty}\|(I-\mathcal{W}_{\tau(n)})\mathcal{B}x_{\tau(n)}\|^{2} = 0. \end{split} \end{eqnarray} | (3.24) |
Employing techniques similar to CASE \mathbb{I} , we have
\begin{eqnarray} \begin{split} \limsup\limits_{n\rightarrow \infty}(\|z\|^{2}-\|x_{\tau(n)+1}\|^{2})\leq0 \end{split} \end{eqnarray} | (3.25) |
and
\begin{eqnarray} \begin{split} \varpi_{\tau(n)+1}^{2} \leq\;&(1-\alpha_{\tau(n)})\varpi_{\tau(n)}^{2} +\alpha_{\tau(n)}(\|z\|^{2}-\|x_{\tau(n)+1}\|^{2}).\\ \end{split} \end{eqnarray} | (3.26) |
Since \varpi_{\tau(n)}\leq\varpi_{\tau(n)+1}, we have
\begin{eqnarray} \begin{split} \varpi^{2}_{\tau(n)} \leq\|z\|^{2}-\|x_{\tau(n)+1}\|^{2}. \end{split} \end{eqnarray} | (3.27) |
By (3.25), we obtain that
\limsup\limits_{n\rightarrow \infty}\varpi_{\tau(n)}\leq0 |
and so
\begin{eqnarray} \lim\limits_{n\rightarrow \infty}\varpi_{\tau(n)} = 0. \end{eqnarray} | (3.28) |
By Eq (3.26), we also obtain
\limsup\limits_{n\rightarrow \infty}\varpi_{\tau(n)+1}\leq \limsup\limits_{n\rightarrow \infty}\varpi_{\tau(n)}. |
In the light of the last inequality and Eq (3.28), we derive that
\lim\limits_{n\rightarrow \infty}\varpi_{\tau(n)+1} = 0. |
Applying Lemma 2.3, we obtain
\varpi_{n}\leq\varpi_{\tau(n)+1}. |
Therefore, we get that \varpi_{n}\rightarrow0 , that is, x_{n}\rightarrow z . This completes the proof.
Algorithm 3.2. Choose an arbitrary initial value x_{1}\in \mathcal{C} . Assume x_{n} has been constructed. Compute
\begin{eqnarray} \begin{split} \;& y_{n} = x_{n}-\tau_{n} \mathcal{B}^\ast(I-\mathcal{T})\mathcal{B}x_{n},\\ \;& z_{n} = \mathcal{P}_{\mathcal{C}}(y_{n}-\varsigma_{n}\mathcal{A}y_{n}),\\ \;&x_{n+1} = \mathcal{P}_{\mathcal{C}}((1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})), \\ \end{split} \end{eqnarray} | (3.29) |
where \{\alpha_{n}\} is a sequence in (0, 1) , \varsigma_{n} is a sequence in (0, \frac{1}{L}) , and \tau_{n} is a sequence in (0, \frac{1}{\|\mathcal{B}\|^{ 2}}) .
Theorem 3.2. If \hat{\Omega}\neq\emptyset and the following conditions are satisfied:
( C_{1} ) lim_{n\rightarrow \infty}\alpha_{n} = 0\ and\ \sum^{\infty}_{n = 0}\alpha_{n} = \infty ;
( C_{2} ) 0 < \liminf_{n\rightarrow \infty}\varsigma_{n}\leq \limsup_{n\rightarrow \infty}\varsigma_{n} < \frac{1}{L} ;
( C_{3} ) 0 < \liminf_{n\rightarrow \infty}\tau_{n}\leq \limsup_{n\rightarrow \infty}\tau_{n} < \frac{1}{\|\mathcal{B}\|^{ 2}} .
Then, the iterative sequence \{x_{n}\} defined by Eq (3.29) strongly converges to the minimum-norm solution \hat{x}( = \mathcal{P}_{\hat{\Omega}}\theta) , where
\hat{\Omega} = \{\hat{x}|\hat{x}\in VI(\mathcal{A},\mathcal{C})\ and \ \mathcal{B}\hat{x}\in Fix(\mathcal{T})\}\neq\emptyset. |
Algorithm 3.3. Choose an arbitrary initial value x_{1}\in \mathcal{C} . Assume x_{n} has been constructed. Compute
\begin{eqnarray} \begin{split} \;& z_{n} = \mathcal{P}_{\mathcal{C}}(x_{n}-\varsigma_{n}\mathcal{A}x_{n}),\\ \;&x_{n+1} = \mathcal{P}_{\mathcal{C}}((1-\alpha_{n})(x_{n}-\varsigma_{n}\mathcal{A}z_{n})), \\ \end{split} \end{eqnarray} | (3.30) |
where \{\alpha_{n}\} is a sequence in (0, 1) and \varsigma_{n} is a sequence in (0, \frac{1}{L}) .
Theorem 3.3. If \hat{\Omega}\neq\emptyset and the following conditions are satisfied:
( C_{1} ) lim_{n\rightarrow \infty}\alpha_{n} = 0\ and\ \sum^{\infty}_{n = 0}\alpha_{n} = \infty ;
( C_{2} ) 0 < \liminf_{n\rightarrow \infty}\varsigma_{n}\leq \limsup_{n\rightarrow \infty}\varsigma_{n} < \frac{1}{L} ;
Then, the iterative sequence \{x_{n}\} defined by Eq (3.30) strongly converges to the minimum-norm solution \hat{x}( = \mathcal{P}_{\Omega}\theta) , where \hat{\Omega} = \{\hat{x}|\hat{x}\in VI(\mathcal{A}, \mathcal{C})\}\neq\emptyset.
In this section, we present some numerical examples to illustrate our main results. The MATLAB codes run in MATLAB version 9.5 (R2018b) on a PC Intel(R) Core(TM)i5-6200 CPU @ 2.30 GHz 2.40 GHz, RAM 8.00 GB. In all examples y -axes shows the value of \|x_{n+1}-x_{n}\| while the x -axis indicates to the number of iterations.
Example 4.1. Let \mathcal{H}_{1} = \mathcal{H}_{2} = R^{n} . The feasible set is defined as:
\mathcal{C}: = \{x\in R^{n}:\|x\|\leq1\}. |
Let G:R^{n}\rightarrow R^{n} is a linear operator defined by:
\mathcal{A}x: = Gx |
for all x\in R^{n} , where G = (g_{ij})_{1\leq i, j \leq n} is a matrix in R^{n\times n} whose terms are given by:
\begin{equation} g_{ij} = \left\{ \begin{aligned} &-1,\ & \mathrm{if}\ j = n+1-i\ \mathrm{and}\ j > i, \\ &1,\ & \mathrm{if}\ j = n+1-i\ \mathrm{and}\ j < i, \\ &0,\ & \mathrm{otherwise}. \\ \end{aligned} \right. \end{equation} | (4.1) |
It is obvious that \mathcal{A} is \|G\| -Lipschitz continuous. By a direct calculation, we also have that \langle \mathcal{A}x, x\rangle = \langle Gx, x\rangle = 0 and so, \mathcal{A} is monotone. Let \mathcal{B} be a matrix in R^{n\times n} which is randomly generated.
Taking cognizance of the difference of the problems handled by Algorithm 3.1 and Algorithm in Tian and Jiang [24], in order to comparing these two algorithms, we make a very small modification to the one in [24] such that it can also solve the problem (1.8). The modified algorithm can be written as follows:
Algorithm 4.1.
\begin{eqnarray} \begin{split} \;& y_{n} = x_{n}-\tau_{n} \mathcal{B}^\ast(I-\mathcal{W}_{n})\mathcal{B}x_{n},\\ \;& z_{n} = \mathcal{P}_{\mathcal{C}}(y_{n}-\varsigma_{n}\mathcal{A}y_{n}),\\ \;&x_{n+1} = \mathcal{P}_{\mathcal{C}}((1-\alpha_{n})(y_{n}-\varsigma_{n}\mathcal{A}z_{n})), \\ \end{split} \end{eqnarray} | (4.2) |
According to the proof of Theorem 3.1, we can easily verify that this modified algorithm works for solving (1.8). The values of control parameters in these two Algorithms are \varsigma_{n} = \frac{1}{2\|G\|} , \tau_{n} = \frac{1}{2\|B\|^{2}} , \alpha_{1} = \frac{1}{2} , \alpha_{n} = \frac{1}{n} (for all n\geq2 ), \lambda_{n} = \frac{1}{n+1} and x_{1} = (1, \cdots, 1)^{T} , and the infinite family of nonexpansive mappings \{\mathcal{T}_{k}\}^{\infty}_{k = 1}:R^{n}\rightarrow R^{n} is defined by:
\mathcal{T}_{k}x: = M_{k}x, |
for all x\in R^{n} , where \{M_{k}\} is a sequence of diagonal matrixes in R^{n\times n} :
\begin{equation} M_{k} = \left[ \begin{array}{ccccc} 1-\frac{1}{k+2} &\ &\ &\ &\ \\ \ &1-\frac{1}{k+2} &\ &\ &\ \\ \ &\ &\ddots &\ &\ \\ \ &\ &\ &1-\frac{1}{k+2} &\ \\ \ &\ &\ &\ &1-\frac{1}{k+3}\\ \end{array} \right ]. \end{equation} | (4.3) |
The numerical results of the Example 4.1 are reported in Table 1 and Figures 1–4 by using the stopping criterion \|x_{n+1}-x_{n}\|\leq10^{-10} .
No. of Iter. | Time | |||
n | Alg. 3.1 | Alg. 4.1 | Alg. 3.1 | Alg. 4.1 |
2 | 120 | 154 | 0.288s | 1.133s |
10 | 156 | 201 | 0.691s | 2.516s |
50 | 157 | 202 | 4.641s | 13.853s |
100 | 157 | 203 | 12.333s | 53.145s |
Example 4.2. Let \mathcal{H}_{1} = \mathcal{H}_{2} = L^{2}([0, 1]) with the inner product:
\langle x,y\rangle = \int^{1}_{0}x(t)y(t)dt |
and the induced norm:
\|x\|: = (\int^{1}_{0}x^{2}(t)dt)^{\frac{1}{2}}. |
The feasible set is defined as:
\mathcal{C}: = \{x\in R^{n}:\|x\|\leq1\}. |
The mapping \mathcal{A}:L^{2}([0, 1])\rightarrow L^{2}([0, 1]) is defined by:
\mathcal{A}x(t): = (1+t)\max\{0,x(t)\} = (1+t)\frac{x(t)+|x(t)|}{2},\ x\in L^{2}([0,1]). |
It is easy to see that
\begin{eqnarray} \begin{split} \langle \mathcal{A}x\;-\mathcal{A}y,x-y\rangle& = \int^{1}_{0}(\mathcal{A}x(t)-\mathcal{A}y(t))(x(t)-y(t))dt \\ \;& = \int^{1}_{0}(1+t)\frac{x(t)-y(t)+|x(t)|-|y(t)|}{2}(x(t)-y(t))dt \\ \;& = \int^{1}_{0}\frac{1}{2}(1+t)((x(t)-y(t))^{2}+(|x(t)|-|y(t)|)(x(t)-y(t)))dt \\ \;&\geq0\\ \end{split} \end{eqnarray} | (4.4) |
and
\begin{eqnarray} \begin{split} \|\mathcal{A}x-\mathcal{A}y\|^{2} \;& = \int^{1}_{0}(\mathcal{A}x(t)-\mathcal{A}y(t))^{2}dt \\ \;& = \int^{1}_{0}(1+t)^{2}\frac{(x(t)-y(t)+|x(t)|-|y(t)|)^{2}}{4}dt \\ \;& = \int^{1}_{0}(1+t)^{2}(x(t)-y(t))^{2}dt \\ \;&\leq4\|x-y\|^{2}.\\ \end{split} \end{eqnarray} | (4.5) |
Therefore, the operator \mathcal{A} is monotone and 2 -Lipschitz continuous. Let \mathcal{W}_{n} = \mathcal{I} (Identity mapping). The values of control parameters for Algorithm 4.1 and Algorithm 3.1 are \varsigma_{n} = \frac{1}{4} , \alpha_{1} = \frac{1}{2} , \alpha_{n} = \frac{1}{n} (for all n\geq2 ), \lambda_{n} = \frac{1}{n+1} and x_{1} = 8t^{2} . It can be seen easily that \{x_{n}\} strongly converges to the zero vector \theta(\in L^{2}([0, 1])). The numerical results of the Example 4.2 are reported in Table 2 and Figures 5 by using the stopping criterion \|x_{n+1}-x_{n}\|\leq\varepsilon = 0.01 .
No. of Iter. | Time | |||
\varepsilon | Alg. 3.1 | Alg. 4.1 | Alg. 3.1 | Alg. 4.1 |
0.01 | 8 | 13 | 0.678s | 79.280s |
Remark 4.1. The numerical results of Example 4.1 and Example 4.2 show that the performance of Algorithm 3.1 is better than Algorithm 4.1 both in CPU time and the number of iterations. Algorithm 3.1 is more effective in both finite and infinite dimensional spaces and especially in conditions involving complex projection calculations, see Tables 1, 2 and Figures 1–5. In Example 4.1, we observe that the number of iterations tends to be stable, while the CPU time increases, as n increasing.
In the present paper, we consider variational inequality problems and split common fixed point problems. We construct an iterative algorithm for solving Eq (1.8) which can be regard as a modification and generalization of Algorithm 1.1 with fewer metric projection operators. Under some mild restrictions, we demonstrate the strong convergence analysis of the presented algorithm. We also give some numerical examples to illustrate our main results. Noticeably, in our article, \mathcal{A} is assumed to a monotone and L -Lipschitz continuous mapping. A natural question arises: how to weaken this assumption?
The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.
This work was supported by National Natural Science Foundation of China under Grant (No. 62103289), National Natural Science Foundation of China (No. 11761074), Project of Jilin Science and Technology Development for Leading Talent of Science and Technology Innovation in Middle and Young and Team Project (No. 20200301053RQ), and Natural Science Foundation of Jilin Province(No. 2020122336JC).
The authors declare no conflict of interest.
[1] |
H. Ahmed, Fractional Euler method: An effective tool for solving fractional differential equtions, J. Egypt. Math. Soc., 26 (2018), 38–43. https://doi.org/10.21608/JOEMS.2018.9460 doi: 10.21608/JOEMS.2018.9460
![]() |
[2] |
D. Chen, I. K. Argyros, The midpoint method for solving nonlinear operator equations in Banach space, Appl. Math. Lett., 5 (1992), 7–9. https://doi.org/10.1016/0893-9659(92)90076-L doi: 10.1016/0893-9659(92)90076-L
![]() |
[3] | J. C. Butcher, Numerical methods for ordinary differential equations, John Wiley & Sons, 2016. http://doi.org/10.1002/9781119121534 |
[4] |
M. Caputo, M. Fabrizio, A new definition of fractional derivative without singular kernel, Progr. Fract. Differ. Appl., 1 (2015), 73–85. http://doi.org/10.12785/pfda/010201 doi: 10.12785/pfda/010201
![]() |
[5] |
C. Carathéodory, Über den Variabilitätsbereich der Koeffizienten von Potenzreihen, die gegebene Werte nicht annehmen, Math. Ann., 64 (1907), 95–115. https://doi.org/10.1007/BF01449883 doi: 10.1007/BF01449883
![]() |
[6] | E. A. Coddington, N. Levinson, T. Teichmann, Theory of ordinary differential equations, New York: McGraw-Hill, 1956. https://doi.org/10.1063/1.3059875 |
[7] | L. Euler, Institutiones calculi integrals, Legare Street Press, 1768. |
[8] | Euler, Foundations of differential calculus, New York: Springer-Verlag, 2000. |
[9] |
B. Ghanbari, K. S. Nisar, Some effective numerical techniques for chaotic systems involving fractal-fractional derivatives with different laws, Front. Phys., 8 (2020), 192. https://doi.org/10.3389/fphy.2020.00192 doi: 10.3389/fphy.2020.00192
![]() |
[10] | E. Hairer, G. Wanner, S. P. Nørsett, Solving ordinary differential equations Ⅰ: Nonstiff problems, Berlin: Springer, 1993. https://doi.org/10.1007/978-3-540-78862-1 |
[11] |
X. Zheng, H. Wang, H. Fu, Well-posedness of fractional differential equations with variable-order Caputo-Fabrizio derivative, Chaos Soliton. Fract., 138 (2020), 109966. https://doi.org/10.1016/j.chaos.2020.109966 doi: 10.1016/j.chaos.2020.109966
![]() |
[12] |
X. Guo, X. Zheng, Variable-order time-fractional diffusion equation with Mittag-Leffler kernel: Regularity analysis and uniqueness of determining variable order, Z. Angew. Math. Phys., 74 (2023), 64. https://doi.org/10.1007/s00033-023-01959-1 doi: 10.1007/s00033-023-01959-1
![]() |
[13] | D. W. Jordan, P. Smith, Nonlinear ordinary differential equations: Problems and solutions: A sourcebook for scientists and engineers, Oxford: Oxford University Press, 2007. |
[14] |
Q. Lai, A. Akgul, C. Li, G. Xu, Ü. Çavuşoğlu, A new chaotic system with multiple attractors: Dynamic analysis, circuit realization and S-Box design, Entropy, 20 (2018), 12. https://doi.org/10.3390/e20010012 doi: 10.3390/e20010012
![]() |
[15] | J. Loustau, Numerical differential equations: Theory and technique, ODE methods, finite differences, finite elements and collocation, World Scientific, 2016. https://doi.org/10.1142/9770 |
[16] | W. Romberg, Vereinfachte numerische integration, Det Kongelige Norske Videnskabers Selskabs, 28 (1955), 30–36. |
[17] |
L. F. Richardson, The approximate arithmetical solution by finite differences of physical problems involving differential equations, with an application to the stresses in a masonry dam, Philos. Trans. Royal Soc. A, 210 (1911), 307–357. https://doi.org/10.1098/rsta.1911.0009 doi: 10.1098/rsta.1911.0009
![]() |
[18] | E. Süli, D. F. Mayers, An introduction to numerical analysis, Cambridge University Press, 2003. https://doi.org/10.1017/CBO9780511801181 |
[19] |
L. W. Neustadt, On the solutions of certain integral-like operator equations, Existence, uniqueness and dependence theorems, Arch. Ration. Mech. Anal., 38 (1970), 131–160. https://doi.org/10.1007/BF00249976 doi: 10.1007/BF00249976
![]() |
No. of Iter. | Time | |||
n | Alg. 3.1 | Alg. 4.1 | Alg. 3.1 | Alg. 4.1 |
2 | 120 | 154 | 0.288s | 1.133s |
10 | 156 | 201 | 0.691s | 2.516s |
50 | 157 | 202 | 4.641s | 13.853s |
100 | 157 | 203 | 12.333s | 53.145s |
No. of Iter. | Time | |||
\varepsilon | Alg. 3.1 | Alg. 4.1 | Alg. 3.1 | Alg. 4.1 |
0.01 | 8 | 13 | 0.678s | 79.280s |