欢迎光临《计算力学学报》官方网站!
基于Armijo准则的自适应稳定转换法
Armijo-based adaptive stability transformation method
投稿时间:2018-05-03  修订日期:2018-06-21
DOI:
中文关键词:  可靠度分析  HL-RF算法  Armijo准则  稳定转换法  高效稳健
英文关键词:reliability analysis  HL-RF algorithm  Armijo rule  stability transformation method  efficient and robust
基金项目:973计划课题(2014CB046506, 2014CB046803);国家自然科学基金(11372061, 11402049, 11602076)项目
作者单位E-mail
李彬 大连理工大学 libindlut2007@126.com 
李刚 大连理工大学 ligang@dlut.edu.cn 
摘要点击次数: 178
全文下载次数: 
中文摘要:
      结构可靠度分析是结构不确定性设计的关键环节,而计算效率和鲁棒性是评估可靠度分析算法性能的两个重要指标。本文首先针对两个已有的一次二阶矩算法(iHL-RF算法和方向性稳定转化法)进行分析,得出iHL-RF算法根据Armijo准则可以自适应调整迭代步长,但计算效率低;方向性稳定转化法根据振荡的方向性可以提高计算效率,但自适应性差。结合两种算法的优点,将Armijo准则用于自适应调整方向性稳定转化法的混沌控制因子,提出了基于Armijo准则的自适应稳定转换法。最后通过四个非线性算例将本文提出的算法与HL-RF、iHL-RF、混沌控制法和方向性稳定转换法等四种算法进行收敛性和计算效率的比较。计算结果表明,相比其他的四种可靠度分析算法,本文提出的算法在求解二维和多维非线性极限状态函数时均具有更好的收敛性和更高的计算效率。
英文摘要:
      Structural reliability analysis is the key part of structural uncertainty design, and efficiency and robustness are two important indexes used to evaluate the performance of reliability analysis methods. At the beginning of this paper, two existing first-order-second-moment methods, named as the iHL-RF method and the directional stability transformation method respectively, are analyzed. It is found that the step length could be adaptively adjusted by the Armijo rule in the iHL-RF method, but the computational efficiency is low. Moreover, the directional stability transformation method could enhance the efficiency based on the direction property of oscillation, but the adaptivity is poor. Combining the advantages of the two algorithms, the Armijo rule could be used to adaptively adjust the chaos control factor of the directional stability transformation method, and then the Armijo-based adaptive stability transformation method is proposed. Finally, based on four nonlinear examples, the convergence and efficiency are compared among five methods including HL-RF, iHL-RF, chaos control method, directional stability transformation method and our proposed method. Numerical results show that the proposed method has better convergence and higher efficiency than other four reliability analysis methods when solving two-dimensional and multi-dimensional nonlinear limit state functions.
  查看/发表评论  下载PDF阅读器
您是第4377148位访问者
版权所有:《计算力学学报》编辑部
本系统由 北京勤云科技发展有限公司设计