分裂可行性问题解集和有限族拟非扩张算子公共不动点集的公共元的迭代算法
Iterative Algorithms of Common Elements for the Set of Solutions of Split Feasibility Problem and the Set of Common Fixed Points of a Finite Family of Quasi-Nonexpansive Operators
通讯作者:
收稿日期: 2024-04-12 修回日期: 2024-08-12
基金资助: |
|
Received: 2024-04-12 Revised: 2024-08-12
Fund supported: |
|
作者简介 About authors
张玉婷,E-mail:
在 Hilbert 空间中, 构造了寻找分裂可行性问题与有限族拟非扩张算子公共不动点问题之公共解的一种新算法. 在适当的条件下, 利用映射的次闭性和投影算子与共轭算子的性质证明了由该算法生成的迭代序列强收敛到分裂可行性问题和不动点问题的公共解, 并给出具体的数值实验验证算法的有效性. 所得结果改进和推广了一些最新文献的相关结果.
关键词:
In real Hilbert spaces, we construct a new algorithm to find a common solution of the split feasibility problem and the fixed points problem involving a finite family of quasi-nonexpansive mappings. Under appropriate conditions, it is proved that the iteration sequence by the algorithm strongly converges to a common solution of the split feasibility problem and the fixed points problem by using the demi-closed principle and properties of projection operators and conjugate operators. The effectiveness of the algorithm is verified by numerical experiments. The results of this paper improve and extend recent some relative results.
Keywords:
本文引用格式
张玉婷, 高兴慧, 彭剑英.
Zhang Yuting, Gao Xinghui, Peng Jianying.
1 引言
设
1994 年, Censor 和 Elfving[1] 在研究 CT 扫描图像重建的过程中首次提出了有限维 Hilbert 空间中的分裂可行性问题. 分裂可行性问题为寻求一点
用 SFP
分裂可行性问题是不动点理论和最优化理论中的一个重要组成部分, 近年来受到了广泛的关注, 它在信号处理、图像重建、模拟调强放射治疗等方面有着很多的应用.
2002 年, Byrne 等人[2]提出了 CQ 算法求解分裂可行性问题 (1.1). 对于任意的
其中,
则我们可以得到
其中,
由于步长的选取
其中,
由于可变步长
其中,
2019 年, Qin 等人[5]提出求解分裂可行性问题和不动点问题的公共解的迭代算法, 并证明了该算法的强收敛性. 具体算法如下
其中,
2024 年, 王元恒等人[6]将算法 (1.3) 中的非扩张算子推广至拟非扩张算子, 并引入了自适应步长和惯性迭代步, 使算法的收敛速度更快. 具体算法如下
他们证明了在适当条件下, 由算法 (1.4) 生成的序列
受上述事实的启发, 本文将算法 (1.4) 中的一个拟非扩张算子推广到有限族拟非扩张算子, 这使得能解决的问题更加广阔, 并提出自适应惯性平行迭代算法, 证明了由该算法生成的序列强收敛到分裂可行性问题与有限族拟非扩张算子不动点问题的公共解, 通过数值实验验证了该算法的收敛速度比算法 (1.4) 的收敛速度更快.
2 预备知识
设
定义2.1[6] 设
(i) 称
(ii) 称
(iii) 称
(iv) 称
(v) 称
定义2.2[6] 令
引理2.1[9] 对于任意的
其中,
回顾度量投影算子
引理2.2[6] 对于
(i)
(ii)
(iii)
由 (i) 易得 (iii) 成立.
引理2.3[10] 令
其中
(i)
(ii)
(iii) 对
则
引理2.4[6] 令
定义2.3[11] 令
那么我们称
引理2.5[12] 令
引理2.6[12] 令
变分不等式问题表述为: 寻求一点
其中,
引理2.7[13] 令
3 主要结果
本文给出如下假设
(B
(B
(B
(B
(B
(B
(B
(B
(B
定义泛函
则有
算法3.1 选取
其中
其中
引理3.1 假设条件 (B
证 首先, 令
任取
根据共轭算子的性质和引理2.2, 由
因此
由 (3.1), (3.3) 式和
根据
将 (3.5) 式代入 (3.4) 式, 可得
由条件 (B
由数学归纳法可得
因此, 序列
定理3.1 假设条件 (B
证 由
由引理 2.6,
其中
根据引理 2.1, 可得
因此
通过 (3.7), (3.9) 和 (3.11) 式, 可得
另一方面, 我们可以得到
其中
将 (3.13)式代入 (3.10) 式, 可得
将 (3.2) 式代入 (3.14) 式, 可得
将 (3.7) 式代入 (3.15) 式, 可得
令
则 (3.12) 和 (3.16) 式可以被表示为
由条件
接下来证明对
选取
由引理 3.1 知
即
于是
结合 (3.18), (3.19) 和 (3.20) 式, 可得
通过上式, 结合
由引理 2.4知
综上,
由
即通过 (3.21) 式可得
综上, 由引理 2.3 知
4 数值实验
例4.1 令
对任意的
令
易知,
在 Alg1 和 Alg2 中, 取初始值 q_0=q_1=[1,1,1,1,1]^T,
考虑
考虑
考虑
考虑
图1
参考文献
A multiprojection algorithm using Bregman projections in a product space
Iterative oblique projection onto convex sets and the split feasibility problem
On variable-step relaxed projection algorithm for variational inequalities
Solving the split feasibility problem without prior knowledge of matrix norms
A viscosity iterative method for a split feasibility problem
两类问题公共解集上的变分不等式解的算法
DOI:10.12386/A20220171
[本文引用: 12]
We study a new algorithm to solve a common solution of the split feasibility problem and the fixed point problem involving quasi-nonexpansive mappings in Hilbert spaces. Based on the common solutions of these two classes of problems, we solve the variational inequality problem. Compared with the predecessors, the self-adaptive technique and the inertial iteration method are added, which can speed up the convergence rate of the iterative sequence generated by our algorithms. At the same time, we extend the involving previous nonexpansive mappings to extensive quasi-nonexpansive mappings. In addition, a strong positive bounded operator is added to the algorithm, which extends the original viscous iterative algorithm to a more general viscous iterative algorithm. The effectiveness of the algorithm is verified by numerical examples.
An algorithm to solve the variational inequality problem based on the common solutions of two classes of problems
DOI:10.12386/A20220171
[本文引用: 12]
We study a new algorithm to solve a common solution of the split feasibility problem and the fixed point problem involving quasi-nonexpansive mappings in Hilbert spaces. Based on the common solutions of these two classes of problems, we solve the variational inequality problem. Compared with the predecessors, the self-adaptive technique and the inertial iteration method are added, which can speed up the convergence rate of the iterative sequence generated by our algorithms. At the same time, we extend the involving previous nonexpansive mappings to extensive quasi-nonexpansive mappings. In addition, a strong positive bounded operator is added to the algorithm, which extends the original viscous iterative algorithm to a more general viscous iterative algorithm. The effectiveness of the algorithm is verified by numerical examples.
Self-Adaptive method and inertial modification for solving the split feasibility problem and fixed-point problem of quasi-nonexpansive mapping
A fixed point method for solving a split feasibility problem in Hilbert spaces
A new iteration method for variational inequalities on the set of common fixed points for a finite family of quasi-pseudocontractions in Hilbert spaces
Solving the variational inequality problem defined on intersection of finite level sets
求解变分不等式和不动点问题的公共元的修正次梯度外梯度算法
Modified Subgradient Extragradient Algorithms for Solving Common Elements of Variational Inequality and Fixed Point Problems
A general iterative method for nonexpansive mappings in Hilbert space
Iterative methods for solving a class of monotone variational inequality problems with applications
/
〈 |
|
〉 |
