基于Armijo非单调线搜索的修正LM方法
A Modified Levenberg-Marquardt Method with an Armijo Nonmonotone Line Search
DOI: 10.12677/AAM.2022.119689, PDF,   
作者: 陈 咪:长沙理工大学数学与统计学院,湖南 长沙
关键词: 非线性方程组LM方法非单调线搜索全局收敛Nonlinear Equations LM Method Nonmonotone Line Search Global Convergence
摘要: 近年来,非线性方程组问题越来越多地出现在科学与工程领域中。Levenberg-Marquardt (LM)方法是解决此问题的有效方法。为了避免信赖域步不可取的情况,文章提出一种基于非单调线搜索技术的修正LM方法,同样保证了算法在局部误差界的条件下达到全局收敛,并在文末附上了相应的数值结果,证明算法是有效的。
Abstract: Recently, systems of nonlinear equations have wide application in fields of science and engineering. The Levenberg-Marquardt method is an effective method to solve this problem. In this paper, we propose a modified Levenberg-Marquardt method by using a nonmonotone line search technique for the nonlinear equations system to avoid the situation where a trust step is not acceptable. The global and cubic convergence of this new method is also preserved under the local error bound con-dition. Finally, some numerical results are reported, which show that the algorithm is efficient.
文章引用:陈咪. 基于Armijo非单调线搜索的修正LM方法[J]. 应用数学进展, 2022, 11(9): 6511-6520. https://doi.org/10.12677/AAM.2022.119689

参考文献

[1] Fan, J.Y. (2012) The Modified Levenberg-Marquardt Method for Nonlinear Equations with Cubic Convergence. Math-ematics of Computation, 81, 447-466. [Google Scholar] [CrossRef
[2] 郭楠, 黄华鹰. 求解非线性方程组的一类非单调修正Levenberg-Marquardt算法[J]. 安徽大学学报(自然科学版), 2016, 40(2): 14-20.
[3] 何叶丹, 马昌凤. 求解非线性方程组的一个修正非单调L-M算法[J]. 福建师范大学学报(自然科学版), 2013, 29(4): 15-22.
[4] Chen, L. and Ma, Y.F. (2020) Shamanskii-Like Levenberg-Marquardt Method with a New Line Search for Systems of Nonlinear Equations. Journal of Systems Science and Complexity, 33, 1694-1707. [Google Scholar] [CrossRef
[5] Zhou, W.J. (2013) On the Convergence of the Modified Leven-berg-Marquardt Method with a Nonmonotone Second Order Armijo Type Line Search. Journal of Computational & Ap-plied Mathematics, 239, 152-161. [Google Scholar] [CrossRef
[6] He, Y.D., Ma, C.F. and Fan, B. (2015) A Corrected Leven-berg-Marquardt Algorithm with a Nonmonotone Line Search for the System of Nonlinear Equations. Applied Mathemat-ics & Computation, 260, 159-169. [Google Scholar] [CrossRef
[7] 周童, 陈亮, 伍珍香. 一种求解非线性方程组的Levenberg-Marquardt方法及其收敛性[J]. 淮北师范大学学报(自然科学版), 2021, 42(1): 1-7.
[8] Amini, K. and Rostami, F. (2016) Three-Steps Modified Levenberg-Marquardt Method with a New Line Search for Systems of Non-linear Equations. Journal of Computational & Applied Mathematics, 300, 30-42. [Google Scholar] [CrossRef
[9] Chen, L. (2016) A Modified Levenberg-Marquardt Method with Line Search for Nonlinear Equations. Computational Optimization & Applications, 65, 753-779. [Google Scholar] [CrossRef
[10] Grippo, L., Lampariello, F. and Lucidi, S. (1986) A Nonmonotone Line Search Technique for Newton’s Method. SIAM Journal on Numerical Analysis, 23, 707-716. [Google Scholar] [CrossRef
[11] Moré, J.J., Garbow, B.S. and Hillstrom, K.E. (1981) Testing Unconstrained Optimization Software. ACM Transform on Mathematical Software, 7, 17-41. [Google Scholar] [CrossRef